# ai
## 5.0.103
### Patch Changes
- Updated dependencies [e609736]
- @ai-sdk/gateway@2.0.16
## 5.0.102
### Patch Changes
- 83b2c4b: fix(ai): prune messages properly when toolCalls set to 'before-last-message'
## 5.0.101
### Patch Changes
- Updated dependencies [c5a42ec]
- @ai-sdk/gateway@2.0.15
## 5.0.100
### Patch Changes
- Updated dependencies [c9520e5]
- Updated dependencies [2382c89]
- @ai-sdk/gateway@2.0.14
## 5.0.99
### Patch Changes
- 15cfac8: feat: export GatewayModelId and use to type LanguageModel
## 5.0.98
### Patch Changes
- Updated dependencies [8c54c71]
- @ai-sdk/gateway@2.0.13
## 5.0.97
### Patch Changes
- Updated dependencies [6465dc3]
- @ai-sdk/gateway@2.0.12
## 5.0.96
### Patch Changes
- c89268c: fix(provider/google): preserve thoughtSignature through tool execution
## 5.0.95
### Patch Changes
- a20e687: feat(provider/gateway): add support for image models
- Updated dependencies [a20e687]
- @ai-sdk/gateway@2.0.11
## 5.0.94
### Patch Changes
- Updated dependencies [646dc54]
- @ai-sdk/gateway@2.0.10
## 5.0.93
### Patch Changes
- Updated dependencies [29fbd2d]
- @ai-sdk/gateway@2.0.9
## 5.0.92
### Patch Changes
- 250edbd: Added finishReason on useChat onFinish callbck
## 5.0.91
### Patch Changes
- Updated dependencies [056c471]
- @ai-sdk/provider-utils@3.0.17
- @ai-sdk/gateway@2.0.8
## 5.0.90
### Patch Changes
- 818b144: fix not catching of empty arrays in validateUIMessage
## 5.0.89
### Patch Changes
- 7cfada6: feat(ai): `chat.addToolResult()` is now `chat.addToolOutput()`
## 5.0.88
### Patch Changes
- Updated dependencies [46dcc02]
- @ai-sdk/gateway@2.0.7
## 5.0.87
### Patch Changes
- 51aa5de: backport: test server
- Updated dependencies [51aa5de]
- @ai-sdk/provider-utils@3.0.16
- @ai-sdk/gateway@2.0.6
## 5.0.86
### Patch Changes
- 0f92046: fix(ui): Don't resend messages for providerExecuted tools in lastAssistantMessageIsCompleteWithToolCalls and lastAssistantMessageIsCompleteWithApprovalResponses
## 5.0.85
### Patch Changes
- Updated dependencies [f2da310]
- @ai-sdk/provider-utils@3.0.15
- @ai-sdk/gateway@2.0.5
## 5.0.84
### Patch Changes
- a017497: fix(ai): do not mutate middleware array argument when wrapping
## 5.0.83
### Patch Changes
- Updated dependencies [872543c]
- @ai-sdk/gateway@2.0.4
## 5.0.82
### Patch Changes
- Updated dependencies [949718b]
- @ai-sdk/provider-utils@3.0.14
- @ai-sdk/gateway@2.0.3
## 5.0.81
### Patch Changes
- 1e05490: fix(ai): do not drop custom headers in HttpChatTransport
- 1e05490: refactor: consolidate header normalization across packages, remove duplicates, preserve custom headers
- Updated dependencies [1e05490]
- @ai-sdk/provider-utils@3.0.13
- @ai-sdk/gateway@2.0.2
## 5.0.80
### Patch Changes
- c7688ec: fix(ai): correct type field in arrayOutputStrategy from 'enum' to 'array'
## 5.0.79
### Patch Changes
- 22ab538: feat(ai): add OAuth for MCP clients + refactor to new package
This change replaces
```ts
import { experimental_createMCPClient } from 'ai';
import { Experimental_StdioMCPTransport } from 'ai/mcp-stdio';
```
with
```ts
import { experimental_createMCPClient } from '@ai-sdk/mcp';
import { Experimental_StdioMCPTransport } from '@ai-sdk/mcp/mcp-stdio';
```
## 5.0.78
### Patch Changes
- Updated dependencies [f1a8734]
- @ai-sdk/gateway@2.0.1
## 5.0.77
### Patch Changes
- 3568ebc: feat(ai): add convertDataPart option to convertToModelMessages
Add optional convertDataPart callback for converting custom data parts (URLs, code files, etc.) to text or file parts that models can process. Fully type-safe using existing UIMessage generics.
## 5.0.76
### Patch Changes
- 87b03d5: feat(ui): add isDataUIPart helper
## 5.0.75
### Patch Changes
- Updated dependencies [76678cd]
- @ai-sdk/gateway@2.0.0
## 5.0.74
### Patch Changes
- Updated dependencies [11a8867]
- @ai-sdk/gateway@1.0.41
## 5.0.73
### Patch Changes
- 734c4c3: fix(ai): add missing export for `LoadSettingError`
## 5.0.72
### Patch Changes
- 1b94257: fix(ui): do not submit automatically when server return with error
## 5.0.71
### Patch Changes
- bc90c3c: feat(ai): add pruneMessages helper function
## 5.0.70
### Patch Changes
- Updated dependencies [f6a9bf3]
- @ai-sdk/gateway@1.0.40
## 5.0.69
### Patch Changes
- 1bacc1f: fix(ai): handle backpressure in `writeToServerResponse`
## 5.0.68
### Patch Changes
- Updated dependencies [17f9872]
- @ai-sdk/provider-utils@3.0.12
- @ai-sdk/gateway@1.0.39
## 5.0.67
### Patch Changes
- Updated dependencies [227ca94]
- @ai-sdk/gateway@1.0.38
## 5.0.66
### Patch Changes
- Updated dependencies [ec5a0a0]
- @ai-sdk/gateway@1.0.37
## 5.0.65
### Patch Changes
- Updated dependencies [638a561]
- @ai-sdk/gateway@1.0.36
## 5.0.64
### Patch Changes
- e33947b: fix(ai): mcp errors to be jsonrpc 2.0 compliant
## 5.0.63
### Patch Changes
- 22609f8: fix(ai): only parse experimental_output in generateText when finishReason is stop
- b552c94: fix(core): Fix image download behavior when the initial model is swapped out during prepareStep
## 5.0.62
### Patch Changes
- 6f0644c: chore: use import \* from zod/v4
- 6f0644c: chore: load zod schemas lazily
- Updated dependencies [6f0644c]
- Updated dependencies [6f0644c]
- @ai-sdk/provider-utils@3.0.11
- @ai-sdk/gateway@1.0.35
## 5.0.61
### Patch Changes
- Updated dependencies [143f65b]
- @ai-sdk/gateway@1.0.34
## 5.0.60
### Patch Changes
- Updated dependencies [cffad55]
- @ai-sdk/gateway@1.0.33
## 5.0.59
### Patch Changes
- Updated dependencies [bcdba23]
- @ai-sdk/gateway@1.0.32
## 5.0.58
### Patch Changes
- Updated dependencies [094685f]
- @ai-sdk/gateway@1.0.31
## 5.0.57
### Patch Changes
- c319849: fixed docs and exported NoSpeechGeneratedError
## 5.0.56
### Patch Changes
- b1d8458: fix(ai): align logic of text-end with reasoning-end
## 5.0.55
### Patch Changes
- 85da29e: fix(ai): remove outdated jsdoc param descriptions
## 5.0.54
### Patch Changes
- bc5ed71: chore: update zod peer depenedency version
- Updated dependencies [bc5ed71]
- @ai-sdk/provider-utils@3.0.10
- @ai-sdk/gateway@1.0.30
## 5.0.53
### Patch Changes
- d335191: Extend addToolResult to support error results
## 5.0.52
### Patch Changes
- c56822d: fix(ai): update `uiMessageChunkSchema` to satisfy the `UIMessageChunk` type
- 930399b: fix(ai): download files when intermediate file cannot be downloaded
- Updated dependencies [7ca78f1]
- @ai-sdk/gateway@1.0.29
## 5.0.51
### Patch Changes
- 27645bb: Export `parseJsonEventStream` and `uiMessageChunkSchema` from "ai" package
- Updated dependencies [322901b]
- @ai-sdk/gateway@1.0.28
## 5.0.50
### Patch Changes
- Updated dependencies [c5f403a]
- @ai-sdk/gateway@1.0.27
## 5.0.49
### Patch Changes
- Updated dependencies [e304478]
- @ai-sdk/gateway@1.0.26
## 5.0.48
### Patch Changes
- Updated dependencies [4d3ff64]
- @ai-sdk/gateway@1.0.25
## 5.0.47
### Patch Changes
- Updated dependencies [c86e0f7]
- Updated dependencies [6bbae01]
- @ai-sdk/gateway@1.0.24
## 5.0.46
### Patch Changes
- c580fbd: Add safeValidateUIMessages utility to validate UI messages without throwing, returning a success/failure result object like Zod’s safeParse
## 5.0.45
### Patch Changes
- 76024fc: fix(ai): fix static tool call and result detection when dynamic is undefined
- 93d8b60: fix(ai): do not filter zero-length text parts that have provider options
- d8eb31f: fix(ai): fix webp image detection from base64
## 5.0.44
### Patch Changes
- Updated dependencies [f49f924]
- @ai-sdk/gateway@1.0.23
## 5.0.43
### Patch Changes
- 0294b58: feat(ai): set `ai`, `@ai-sdk/provider-utils`, and runtime in `user-agent` header
- Updated dependencies [0294b58]
- @ai-sdk/provider-utils@3.0.9
- @ai-sdk/gateway@1.0.22
## 5.0.42
### Patch Changes
- de5c066: fix(ai): forwarded providerExecuted flag in validateUIMessages
## 5.0.41
### Patch Changes
- cd91e4b: fix(ai): use correct type for reasoning outputs
## 5.0.40
### Patch Changes
- Updated dependencies [4ee3719]
- @ai-sdk/gateway@1.0.21
## 5.0.39
### Patch Changes
- a0a725f: feat (ai): export createGateway
## 5.0.38
### Patch Changes
- Updated dependencies [350a328]
- @ai-sdk/gateway@1.0.20
## 5.0.37
### Patch Changes
- d6785d7: feat (ai): add tool and agent helpers
## 5.0.36
### Patch Changes
- ccc2ded: feat (ai): export gateway provider
## 5.0.35
### Patch Changes
- 99c946a: export missing type
## 5.0.34
### Patch Changes
- Updated dependencies [034287f]
- Updated dependencies [dee1afe]
- @ai-sdk/gateway@1.0.19
## 5.0.33
### Patch Changes
- Updated dependencies [5d59a8c]
- @ai-sdk/gateway@1.0.18
## 5.0.32
### Patch Changes
- Updated dependencies [b6005cd]
- @ai-sdk/gateway@1.0.17
## 5.0.31
### Patch Changes
- Updated dependencies [99964ed]
- @ai-sdk/provider-utils@3.0.8
- @ai-sdk/gateway@1.0.16
## 5.0.30
### Patch Changes
- 7fcc6be: feat(ai): throw InvalidArgumentError when messages is not provided
## 5.0.29
### Patch Changes
- e0e9449: feat(ui): sent isAbort, isDisconnect, isError in useChat onFinish callback
## 5.0.28
### Patch Changes
- 4b81e7d: fix(ai): remove vitest dependency from test exports
- d68a4f2: feat(ai): log warnings
## 5.0.27
### Patch Changes
- ca40fac: feat(ai): support custom download functions (experimental)
## 5.0.26
### Patch Changes
- 33cf848: feat(ai): pass messages to `useChat({ onFinish })`
- Updated dependencies [980633d]
- Updated dependencies [1c5b88d]
- @ai-sdk/gateway@1.0.15
## 5.0.25
### Patch Changes
- ca65923: fix(ai): remove use of `expect()` from production code
- Updated dependencies [886e7cd]
- @ai-sdk/provider-utils@3.0.7
- @ai-sdk/gateway@1.0.14
## 5.0.24
### Patch Changes
- f8f3682: fix: call onFinish when stream is cancelled in toUIMessageStream
Previously, onFinish was only called on normal stream completion. Now it's also called when the reader is cancelled (e.g., browser close, navigation), ensuring partial messages are persisted.
- Updated dependencies [1b5a3d3]
- Updated dependencies [c9994f9]
- @ai-sdk/provider-utils@3.0.6
- @ai-sdk/gateway@1.0.13
## 5.0.23
### Patch Changes
- 5099b3d: fix(ai): make `chat.addToolResult()` compatible with dynamic tool calls
- 7a2bf8d: fix(ai): support loop breaking behavior in async iterable stream
- Updated dependencies [50e2029]
- Updated dependencies [b8478f0]
- @ai-sdk/gateway@1.0.12
## 5.0.22
### Patch Changes
- Updated dependencies [926259f]
- Updated dependencies [c000f96]
- @ai-sdk/gateway@1.0.11
## 5.0.21
### Patch Changes
- 581abea: fix(ai): call abort callback when stream is aborted during tool execution
- 3c178ec: feat(ai): improved type checking for prompt/messages input
- Updated dependencies [0857788]
- @ai-sdk/provider-utils@3.0.5
- @ai-sdk/gateway@1.0.10
## 5.0.20
### Patch Changes
- 8a87693: fix(ai) Make sure warnings promise in streamObject is resolved and properly collects and passes warnings
## 5.0.19
### Patch Changes
- 8da6e9c: fix(ai): use parsed tool input if possible when validation fails
## 5.0.18
### Patch Changes
- Updated dependencies [8b96f99]
- @ai-sdk/gateway@1.0.9
## 5.0.17
### Patch Changes
- 4176ecb: feat(ai): add reasoning text to generateObject result
- 20f23f9: feat(ai): export LanguageModelMiddleware type
## 5.0.16
### Patch Changes
- Updated dependencies [68751f9]
- @ai-sdk/provider-utils@3.0.4
- @ai-sdk/gateway@1.0.8
## 5.0.15
### Patch Changes
- ca4f68f: feat(ai): add validateUIMessages function
- Updated dependencies [28a4006]
- @ai-sdk/gateway@1.0.7
## 5.0.14
### Patch Changes
- 7729e32: fix(ai): expand mp3 detection to support all mpeg frame headers
## 5.0.13
### Patch Changes
- a7b2e66: Added providerOptions to agent stream and generate calls
- 9bed210: ### `extractReasoningMiddleware()`: delay sending `text-start` chunk to prevent rendering final text before reasoning
When wrapping a text stream in `extractReasoningMiddleware()`, delay queing the `text-start` chunk until either `reasoning-start` chunk was queued or the first `text-delta` chunk is about to be queued, whichever comes first.
https://github.com/vercel/ai/pull/8036
## 5.0.12
### Patch Changes
- Updated dependencies [eefa730]
- Updated dependencies [034e229]
- Updated dependencies [f25040d]
- @ai-sdk/gateway@1.0.6
- @ai-sdk/provider-utils@3.0.3
## 5.0.11
### Patch Changes
- 38ac190: feat(ai): preliminary tool results
- e3a63cb: fix(ai): streamText promises reject when stream has errors
- Updated dependencies [38ac190]
- Updated dependencies [cf7b2ad]
- @ai-sdk/provider-utils@3.0.2
- @ai-sdk/gateway@1.0.5
## 5.0.10
### Patch Changes
- 63a5dc5: fix(ai): convert user message text/file part provider metadata in convertToModelMessages
## 5.0.9
### Patch Changes
- afd5c2a: fix(ai): preserve filename for file parts in convertToModelMessages
## 5.0.8
### Patch Changes
- Updated dependencies [35f93ce]
- @ai-sdk/gateway@1.0.4
## 5.0.7
### Patch Changes
- 8e72304: fix (ai): handle invalid tool calls
## 5.0.6
### Patch Changes
- d983eee: feat(ai): allow passing model string for embeddings
## 5.0.5
### Patch Changes
- Updated dependencies [893aed6]
- @ai-sdk/gateway@1.0.3
## 5.0.4
### Patch Changes
- Updated dependencies [444df49]
- @ai-sdk/gateway@1.0.2
## 5.0.3
### Patch Changes
- 90d212f: feat (ai): add experimental tool call context
- Updated dependencies [028fb9c]
- Updated dependencies [90d212f]
- Updated dependencies [6331826]
- @ai-sdk/gateway@1.0.1
- @ai-sdk/provider-utils@3.0.1
## 5.0.2
### Patch Changes
- 401d73e: fix (ai): support dynamic tool calls in lastAssistantMessageIsCompleteWithToolCalls
- 69fde99: Set status to ready when reconnect stream is null
## 5.0.1
### Patch Changes
- 4d0c108: fix(ai/ui): convert provider metadata for system messages to model messages
## 5.0.0
### Major Changes
- e1cbf8a: chore(@ai-sdk/rsc): extract to separate package
- a847c3e: chore: rename reasoning to reasoningText etc
- 13fef90: chore (ai): remove automatic conversion of UI messages to model messages
- d964901: - remove setting temperature to `0` by default
- remove `null` option from `DefaultSettingsMiddleware`
- remove setting defaults for `temperature` and `stopSequences` in `ai` to enable middleware changes
- 0a710d8: feat (ui): typed tool parts in ui messages
- 9ad0484: feat (ai): automatic tool execution error handling
- 63f9e9b: chore (provider,ai): tools have input/output instead of args,result
- ab7ccef: chore (ai): change source ui message parts to source-url
- d5f588f: AI SDK 5
- ec78cdc: chore (ai): remove "data" UIMessage role
- 6a83f7d: refactoring (ai): restructure message metadata transfer
- db345da: chore (ai): remove exports of internal ui functions
- 496bbc1: chore (ui): inline/remove ChatRequest type
- 72d7d72: chore (ai): stable activeTools
- 40acf9b: feat (ui): introduce ChatStore and ChatTransport
- 98f25e5: chore (ui): remove managed chat inputs
- 2d03e19: chore (ai): remove StreamCallbacks.onCompletion
- da70d79: chore (ai): remove getUIText helper
- c60f895: chore (ai): remove useChat keepLastMessageOnError
- 0560977: chore (ai): improve consistency of generate text result, stream text result, and step result
- 9477ebb: chore (ui): remove useAssistant hook (**breaking change**)
- 1f55c21: chore (ai): send reasoning to the client by default
- e7dc6c7: chore (ai): remove onResponse callback
- 8b86e99: chore (ai): replace `Message` with `UIMessage`
- 04d5063: chore (ai): rename default provider global to AI_SDK_DEFAULT_PROVIDER
- 319b989: chore (ai): remove content from ui messages
- 14c9410: chore: refactor file towards source pattern (spec)
- a34eb39: chore (ai): remove `data` and `allowEmptySubmit` from `ChatRequestOptions`
- f04fb4a: chore (ai): replace useChat attachments with file ui parts
- f7e8bf4: chore (ai): flatten ui message stream parts
- 257224b: chore (ai): separate TextStreamChatTransport
- fd1924b: chore (ai): remove redundant `mimeType` property
- 2524fc7: chore (ai): remove ui message toolInvocations property
- 6fba4c7: chore (ai): remove deprecated experimental_providerMetadata
- b4b4bb2: chore (ui): rename experimental_resume to resumeStream
- 441d042: chore (ui): data stream protocol v2 with SSEs
- ef256ed: chore (ai): refactor and use chatstore in svelte
- 516be5b: ### Move Image Model Settings into generate options
Image Models no longer have settings. Instead, `maxImagesPerCall` can be passed directly to `generateImage()`. All other image settings can be passed to `providerOptions[provider]`.
Before
```js
await generateImage({
model: luma.image('photon-flash-1', {
maxImagesPerCall: 5,
pollIntervalMillis: 500,
}),
prompt,
n: 10,
});
```
After
```js
await generateImage({
model: luma.image('photon-flash-1'),
prompt,
n: 10,
maxImagesPerCall: 5,
providerOptions: {
luma: { pollIntervalMillis: 5 },
},
});
```
Pull Request: https://github.com/vercel/ai/pull/6180
- a662dea: chore (ai): remove sendExtraMessageFields
- d884051: feat (ai): simplify default provider setup
- e8324c5: feat (ai): add args callbacks to tools
- fafc3f2: chore (ai): change file to parts to use urls instead of data
- 1ed0287: chore (ai): stable sendStart/sendFinish options
- c7710a9: chore (ai): rename DataStreamToSSETransformStream to JsonToSseTransformStream
- bfbfc4c: feat (ai): streamText/generateText: totalUsage contains usage for all steps. usage is for a single step.
- 9ae327d: chore (ui): replace chat store concept with chat instances
- 9315076: chore (ai): rename continueUntil to stopWhen. Rename maxSteps stop condition to stepCountIs.
- 247ee0c: chore (ai): remove steps from tool invocation ui parts
- 109c0ac: chore (ai): rename id to chatId (in post request, resume request, and useChat)
- 954aa73: feat (ui): extended regenerate support
- 33eb499: feat (ai): inject message id in createUIMessageStream
- 901df02: feat (ui): use UI_MESSAGE generic
- 4892798: chore (ai): always stream tool calls
- c25cbce: feat (ai): use console.error as default error handler for streamText and streamObject
- b33ed7a: chore (ai): rename DataStream* to UIMessage*
- ed675de: feat (ai): add ui data parts
- 7bb58d4: chore (ai): restructure prepareRequest
- ea7a7c9: feat (ui): UI message metadata
- 0463011: fix (ai): update source url stream part
- dcc549b: remove StreamTextResult.mergeIntoDataStream method
rename DataStreamOptions.getErrorMessage to onError
add pipeTextStreamToResponse function
add createTextStreamResponse function
change createDataStreamResponse function to accept a DataStream and not a DataStreamWriter
change pipeDataStreamToResponse function to accept a DataStream and not a DataStreamWriter
change pipeDataStreamToResponse function to have a single parameter
- 35fc02c: chore (ui): rename RequestOptions to CompletionRequestOptions
- 64f6d64: feat (ai): replace maxSteps with continueUntil (generateText)
- 175b868: chore (ai): rename reasoning UI parts 'reasoning' property to 'text'
- 60e2c56: feat (ai): restructure chat transports
- 765f1cd: chore (ai): remove deprecated useChat isLoading helper
- cb2b53a: chore (ai): refactor header preparation
- e244a78: chore (ai): remove StreamData and mergeStreams
- d306260: feat (ai): replace maxSteps with continueUntil (streamText)
- 4bfe9ec: chore (ai): remove ui message reasoning property
- 1766ede: chore: rename maxTokens to maxOutputTokens
- 2877a74: chore (ai): remove ui message data property
- 1409e13: chore (ai): remove experimental continueSteps
- b32e192: chore (ai): rename reasoning to reasoningText, rename reasoningDetails to reasoning (streamText, generateText)
- 92cb0a2: chore (ai): rename CoreMessage to ModelMessage
- 2b637d6: chore (ai): rename UIMessageStreamPart to UIMessageChunk
### Minor Changes
- b7eae2d: feat (core): Add finishReason field to NoObjectGeneratedError
- bcea599: feat (ai): add content to generateText result
- 48d675a: feat (ai): add content to streamText result
- c9ad635: feat (ai): add filename to file ui parts
### Patch Changes
- a571d6e: chore(provider-utils): move ToolResultContent to provider-utils
- de2d2ab: feat(ai): add provider and provider registry middleware functionality
- c22ad54: feat(smooth-stream): chunking callbacks
- d88455d: feat (ai): expose http chat transport type
- e7fcc86: feat (ai): introduce dynamic tools
- da1e6f0: feat (ui): add generics to ui message stream parts
- 48378b9: fix (ai): send null as tool output when tools return undefined
- 5d1e3ba: chore (ai): remove provider re-exports
- 93d53a1: chore (ai): remove cli
- e90d45d: chore (rsc): move HANGING_STREAM_WARNING_TIME constant into @ai-sdk/rsc package
- b32c141: feat (ai): add array support to stopWhen
- bc3109f: chore (ai): push stream-callbacks into langchain/llamaindex adapters
- 0d9583c: fix (ai): use user-provided media type when available
- 38ae5cc: feat (ai): export InferUIMessageChunk type
- 10b21eb: feat(cli): add ai command line interface
- 9e40cbe: Allow destructuring output and errorText on `ToolUIPart` type
- 6909543: feat (ai): support system parameter in Agent constructor
- 86cfc72: feat (ai): add ignoreIncompleteToolCalls option to convertToModelMessages
- 377bbcf: fix (ui): tool input can be undefined during input-streaming
- d8aeaef: feat(providers/fal): add transcribe
- ae77a99: chore (ai): rename text and reasoning chunks in streamText fullstream
- 4fef487: feat: support for zod v4 for schema validation
All these methods now accept both a zod v4 and zod v3 schemas for validation:
- `generateObject()`
- `streamObject()`
- `generateText()`
- `experimental_useObject()` from `@ai-sdk/react`
- `streamUI()` from `@ai-sdk/rsc`
- b1e3abd: feat (ai): expose ui message stream headers
- 4f3e637: fix (ui): avoid caching globalThis.fetch in case it is patched by other libraries
- 14cb3be: chore(providers/llamaindex): extract to separate package
- 1f6ce57: feat (ai): infer tool call types in the `onToolCall` callback
- 16ccfb2: feat (ai): add readUIMessageStream helper
- 225f087: fix (ai/mcp): prevent mutation of customEnv
- ce1d1f3: feat (ai): export mock image, speech, and transcription models
- fc0380b: feat (ui): resolvable header, body, credentials in http chat transport
- 6622441: feat (ai): add static/dynamic toolCalls/toolResults helpers
- 4048ce3: fix (ai): add tests and examples for openai responses
- 6c42e56: feat (ai): validate ui stream data chunks
- bedb239: chore (ai): make ui stream parts value optional when it's not required
- 9b4d074: feat(streamObject): add enum support
- c8fce91: feat (ai): add experimental Agent abstraction
- 655cf3c: feat (ui): add onFinish to createUIMessageStream
- 3e10408: fix(utils/detect-mimetype): add support for detecting id3 tags
- d5ae088: feat (ui): add sendAutomaticallyWhen to Chat
- ced8eee: feat(ai): re-export zodSchema from main package
- c040e2f: fix (ui): inject generated response message id
- d3960e3: selectTelemetryAttributes more robustness
- faea29f: fix (provider/openai): multi-step reasoning with text
- 66af894: fix (ai): respect content order in toResponseMessages
- 332167b: chore (ai): move maxSteps into UseChatOptions
- 6b1c55c: feat (ai): introduce GLOBAL_DEFAULT_PROVIDER
- 5a975a4: feat (ui): update Chat tool result submission
- 507ac1d: fix (ui/react): update messages immediately with the submitted user message
- a166433: feat: add transcription with experimental_transcribe
- 26735b5: chore(embedding-model): add v2 interface
- c93a8bc: chore(ai): export AsyncIterableStream type from async-iterable-stream module
- 0d2c085: feat (ai): support string model ids through gateway
- 2b9bbcd: feat (ai): improve prompt validation error message
- a8c8bd5: feat(embed-many): respect supportsParallelCalls & concurrency
- 75c3396: fix (ai): handle errors in 2nd streamText doStream call
- cb9c9e4: remove deprecated `experimental_wrapLanguageModel`
- 9bf7291: chore(providers/openai): enable structuredOutputs by default & switch to provider option
- 9b0da33: fix (ai): do not send id with start unless specified
- 28ad69e: fix(react-native): support experimental_attachments without FileList global
- 0b78e17: chore(ai/generateObject): simplify function signature
- 20398f2: feat: ai sdk cli documentation + adjusted default model
- 66962ed: fix(packages): export node10 compatible types
- b71fe8d: fix(ai): remove jsondiffpatch dependency
- 7827a49: fix (ai/core): refactor `toResponseMessages` to filter out empty string/content
- bd8a36c: feat(embedding-model-v2/embedMany): add response body field
- d9209ca: fix (image-model): `specificationVersion: v1` -> `v2`
- b346545: feat (ai): add data ui part schemas
- 05d2819: feat: allow zod 4.x as peer dependency
- f2b041e: Fix custom `fetch` in HttpChatTransport
- 2a62513: Fix error thrown when emptying messages in onError or onFinish
- 143c55b: feat (ai): export Chat callback types
- 9301f86: refactor (image-model): rename `ImageModelV1` to `ImageModelV2`
- 904fa5e: feat (ai/core): add terminateOnError option to readUIMessageStream
- 0a87932: core (ai): change transcription model mimeType to mediaType
- 1675396: fix: avoid job executor deadlock when adding tool result
- 51f497d: feat (ai): step input message modification in prepareStep
- cee64b2: fix(otel): change back toolCall attributes of input/output back to args/result for compatibility
- f04ffe4: feat (ui): add onData callback to Chat
- bc24722: feat (ai): Add finishReason as a promise on StreamObjectResult to match StreamTextResult
- b6f9f3c: remove deprecated `CoreTool*` types
- 8aa9e20: feat: add speech with experimental_generateSpeech
- 4617fab: chore(embedding-models): remove remaining settings
- 8255639: ### Fix use with Google APIs + zod v4's `.literal()` schema
Before [zod@3.25.49](https://github.com/colinhacks/zod/releases/tag/v3.25.49), requests to Google's APIs failed due to a missing `type` in the provided schema. The problem has been resolved for the `ai` SDK by bumping our `zod` peer dependencies to `^3.25.49`.
pull request: https://github.com/vercel/ai/pull/6609
- f81c720: chore(ai): bundle dependencies in CLI binary
- cf9af6e: feat (ai): allow sync prepareStep
- ee38081: Add support for audio/webm to detect-media-type
- 2e4f9e4: feat (ai): improved error messages when using gateway
- 3e3b9df: fix (ai/mcp): better support for zero-argument MCP tools
- cda32ba: fix (ai): send `start` part in correct position in stream (streamText)
- 48a7606: feat (ai): support changing the system prompt in prepareSteps
- cb68df0: feat: add transcription and speech model support to provider registry
- db64cbe: fix (provider/openai): multi-step reasoning with tool calls
- 97c35c0: feat (ui): transient data parts
- 26695a3: feat (ui): add state for text and reasoning ui message parts
- 90ac328: fix (ui): tool part metadata support in ui messages
- 60132dd: fixed date formatting for updated mcp protocol version
- 4a1e0c8: fix(ai-cli): fix bundling and improve authentication error handling
- c6b64a7: feat (ai): allow async prepareRequest on HttpChatTransport
- fccf75c: update mcp protocol version
- 9121250: Expose provider metadata as an attribute on exported OTEL spans
- ea27cc6: chore (ai): use JSONValue definition from provider
- 90ca2b9: feat(ai): Record tool call errors on tool call spans recorded in `generateText` and `streamText`.
- 50f0362: fix (ai): fix experimental sendStart/sendFinish options in streamText
- 825e8d7: release alpha.5
- 7d97ab6: release alpha.4
- 0ff02bb: chore(provider-utils): move over jsonSchema
- 4f3776c: feat (ai): add InferUITools helper
- 9338f3e: fix (ai): throw error for v1 models
- 92c8e66: fix(ai/core): properly handle custom separator in provider registry
- 53569b8: feat (ai): add experimental repairText function to streamObject
- 82aa95d: fix (ai): merge data ui stream parts correctly
- e7d2ce3: feat: provider-executed tools
- add5ac1: feat (ai): make streamText toUIMessageStream async iterable
- 37a916d: feat (ai): add prepareSteps to streamText
- 30ac566: fix (ui): text message metadata support in ui messages
- 8026705: fix (core): send buffered text in smooth stream when stream parts change
- 9bd5ab5: feat (provider): add providerMetadata to ImageModelV2 interface (#5977)
The `experimental_generateImage` method from the `ai` package now returnes revised prompts for OpenAI's image models.
```js
const prompt = 'Santa Claus driving a Cadillac';
const { providerMetadata } = await experimental_generateImage({
model: openai.image('dall-e-3'),
prompt,
});
const revisedPrompt = providerMetadata.openai.images[0]?.revisedPrompt;
console.log({
prompt,
revisedPrompt,
});
```
- ec5933d: chore (ai/mcp): add `assertCapability` method to experimental MCP client
- 09f41ac: fix (ui): add message metadata in Chat.sendMessage
- ff1c81a: feat (ai): add streamText onAbort callback
- af1d5a5: fix(ai): Unexpected reasoning-start event in extract reasoning middleware
- cb3b9c9: fix (ai): catch errors in ui message stream
- 86293e5: fix (ai): use correct generateMessageId in streamText toUIMessageStream
- d1a034f: feature: using Zod 4 for internal stuff
- fd65bc6: chore(embedding-model-v2): rename rawResponse to response
- d92b9a8: fix(ai): add support for MCP protocol version 2025-06-18
- 102b066: fix (ai): fix invalid fetch call
- 142576e: feat (ui): support message replacement in chat via messageId param on sendMessage
- 84343eb: fix (ui): call sendAutomaticallyWhen with updated messages
- a76a62b: feat (ai): add experimental prepareStep callback to generateText
- 89ba235: fix (ui): support tool names with dash
- 8e31d46: feat (ai): export SourceDocumentUIPart
- bd398e4: fix (core): improve error handling in streamText's consumeStream method
- 88a8ee5: fix (ai): support abort during retry waits
- 205077b: fix: improve Zod compatibility
- d91b50d: chore(ui-utils): merge into ai package
- e4c8647: feat (ui): allow asynchronous onFinish in createUIMessageStream
- c808e4d: fix (ui): do not send changing assistant message ids when onFinish is provided
- e862b5b: feat (ai): allow sync tool.execute
- 395c85e: feat (ai): add consumeSseStream option to UI message stream responses
- 5bdff05: Removed deprecated `options.throwErrorForEmptyVectors` from `cosineSimilarity()`. Since `throwErrorForEmptyVectors` was the only option the entire `options` argument was removed.
```diff
- cosineSimilarity(vector1, vector2, options)
+cosineSimilarity(vector1, vector2)
```
- 13b4f46: feat (ai): export experimental MCPClient and MCPClientConfig interfaces
- a4f3007: chore: remove ai/react
- 8e64e9c: feat (ai): allow using provider default temperature by specifying null
- b983b51: feat (ai): support model message array in prompt
- 56c232b: fix (ai): remove outdated sendStart jsdoc
- 7324c21: fix (ai/telemetry): Avoid JSON.stringify on Uint8Arrays for telemetry
- f10304b: feat(tool-calling): don't require the user to have to pass parameters
- dd5fd43: feat (ai): support dynamic tools in Chat onToolCall
- a753b3a: feat (provider/anthropic): cache control for tools
- 383cbfa: feat (ai): add isAborted to onFinish callback for ui message streams
- 27deb4d: feat (provider/gateway): Add providerMetadata to embeddings response
- 5f2b3d4: chore (ai): stable prepareStep
- 4c8f834: feat: automatically respect rate limit headers in retry logic
Added automatic support for respecting rate limit headers (`retry-after-ms` and `retry-after`) in the SDK's retry logic. When these headers are present and contain reasonable values (0-60 seconds), the retry mechanism will use the server-specified delay instead of exponential backoff. This matches the behavior of Anthropic and OpenAI client SDKs and improves rate limit handling without requiring any API changes.
- f2c7f19: feat (ui): add Chat.clearError()
- 7bd025b: fix (ai): fix sync tool execute with streamText
- Updated dependencies [a571d6e]
- Updated dependencies [742b7be]
- Updated dependencies [9e16bfd]
- Updated dependencies [e7fcc86]
- Updated dependencies [0477a13]
- Updated dependencies [7cddb72]
- Updated dependencies [ccce59b]
- Updated dependencies [e2b9e4b]
- Updated dependencies [95857aa]
- Updated dependencies [45c1ea2]
- Updated dependencies [6f6bb89]
- Updated dependencies [26b6dd0]
- Updated dependencies [060370c]
- Updated dependencies [dc714f3]
- Updated dependencies [b5da06a]
- Updated dependencies [d1a1aa1]
- Updated dependencies [63f9e9b]
- Updated dependencies [5d142ab]
- Updated dependencies [d5f588f]
- Updated dependencies [30ab1de]
- Updated dependencies [e025824]
- Updated dependencies [0571b98]
- Updated dependencies [b6b43c7]
- Updated dependencies [4fef487]
- Updated dependencies [48d257a]
- Updated dependencies [0c0c0b3]
- Updated dependencies [0d2c085]
- Updated dependencies [40acf9b]
- Updated dependencies [9222aeb]
- Updated dependencies [e2aceaf]
- Updated dependencies [411e483]
- Updated dependencies [97fedf9]
- Updated dependencies [8ba77a7]
- Updated dependencies [c91586a]
- Updated dependencies [7b3ae3f]
- Updated dependencies [a166433]
- Updated dependencies [3cbcbb7]
- Updated dependencies [26735b5]
- Updated dependencies [443d8ec]
- Updated dependencies [a8c8bd5]
- Updated dependencies [abf9a79]
- Updated dependencies [14c9410]
- Updated dependencies [e86be6f]
- Updated dependencies [9bf7291]
- Updated dependencies [2e13791]
- Updated dependencies [9f95b35]
- Updated dependencies [66962ed]
- Updated dependencies [fedb55e]
- Updated dependencies [0d06df6]
- Updated dependencies [472524a]
- Updated dependencies [dd3ff01]
- Updated dependencies [6c2c708]
- Updated dependencies [d9c98f4]
- Updated dependencies [05d2819]
- Updated dependencies [9301f86]
- Updated dependencies [0a87932]
- Updated dependencies [c4a2fec]
- Updated dependencies [957b739]
- Updated dependencies [721775e]
- Updated dependencies [70ebead]
- Updated dependencies [79457bd]
- Updated dependencies [a3f768e]
- Updated dependencies [f3639fa]
- Updated dependencies [7435eb5]
- Updated dependencies [8aa9e20]
- Updated dependencies [4617fab]
- Updated dependencies [ac34802]
- Updated dependencies [0054544]
- Updated dependencies [cb68df0]
- Updated dependencies [ad80501]
- Updated dependencies [8bd3624]
- Updated dependencies [68ecf2f]
- Updated dependencies [9e9c809]
- Updated dependencies [32831c6]
- Updated dependencies [6dc848c]
- Updated dependencies [6b98118]
- Updated dependencies [d0f9495]
- Updated dependencies [c145d62]
- Updated dependencies [63d791d]
- Updated dependencies [87b828f]
- Updated dependencies [3f2f00c]
- Updated dependencies [bfdca8d]
- Updated dependencies [0ff02bb]
- Updated dependencies [7979f7f]
- Updated dependencies [f77bc38]
- Updated dependencies [989ac75]
- Updated dependencies [39a4fab]
- Updated dependencies [44f4aba]
- Updated dependencies [7742ba3]
- Updated dependencies [9bd5ab5]
- Updated dependencies [57edfcb]
- Updated dependencies [faf8446]
- Updated dependencies [c190907]
- Updated dependencies [7ea4132]
- Updated dependencies [d1a034f]
- Updated dependencies [d454e4b]
- Updated dependencies [5c56081]
- Updated dependencies [fd65bc6]
- Updated dependencies [cf1e00e]
- Updated dependencies [023ba40]
- Updated dependencies [cc21603]
- Updated dependencies [ea7a7c9]
- Updated dependencies [26535e0]
- Updated dependencies [e030615]
- Updated dependencies [5e57fae]
- Updated dependencies [393138b]
- Updated dependencies [c57e248]
- Updated dependencies [88a8ee5]
- Updated dependencies [41fa418]
- Updated dependencies [205077b]
- Updated dependencies [71f938d]
- Updated dependencies [e001ea1]
- Updated dependencies [3795467]
- Updated dependencies [28a5ed5]
- Updated dependencies [7182d14]
- Updated dependencies [c1e6647]
- Updated dependencies [1766ede]
- Updated dependencies [811dff3]
- Updated dependencies [f10304b]
- Updated dependencies [dd5fd43]
- Updated dependencies [33f4a6a]
- Updated dependencies [383cbfa]
- Updated dependencies [27deb4d]
- Updated dependencies [c4df419]
- @ai-sdk/provider-utils@3.0.0
- @ai-sdk/provider@2.0.0
- @ai-sdk/gateway@1.0.0
## 5.0.0-beta.34
### Patch Changes
- 53569b8: feat (ai): add experimental repairText function to streamObject
- 88a8ee5: fix (ai): support abort during retry waits
- f2c7f19: feat (ui): add Chat.clearError()
- Updated dependencies [721775e]
- Updated dependencies [88a8ee5]
- @ai-sdk/gateway@1.0.0-beta.19
- @ai-sdk/provider-utils@3.0.0-beta.10
## 5.0.0-beta.33
### Patch Changes
- 48378b9: fix (ai): send null as tool output when tools return undefined
- 93d53a1: chore (ai): remove cli
- 27deb4d: feat (provider/gateway): Add providerMetadata to embeddings response
- Updated dependencies [27deb4d]
- @ai-sdk/gateway@1.0.0-beta.18
- @ai-sdk/provider@2.0.0-beta.2
- @ai-sdk/provider-utils@3.0.0-beta.9
## 5.0.0-beta.32
### Patch Changes
- bc24722: feat (ai): Add finishReason as a promise on StreamObjectResult to match StreamTextResult
- 13b4f46: feat (ai): export experimental MCPClient and MCPClientConfig interfaces
- 56c232b: fix (ai): remove outdated sendStart jsdoc
## 5.0.0-beta.31
### Patch Changes
- 6622441: feat (ai): add static/dynamic toolCalls/toolResults helpers
- ced8eee: feat(ai): re-export zodSchema from main package
- cee64b2: fix(otel): change back toolCall attributes of input/output back to args/result for compatibility
- ee38081: Add support for audio/webm to detect-media-type
- dd5fd43: feat (ai): support dynamic tools in Chat onToolCall
- Updated dependencies [dd5fd43]
- @ai-sdk/provider-utils@3.0.0-beta.8
- @ai-sdk/gateway@1.0.0-beta.17
## 5.0.0-beta.30
### Patch Changes
- Updated dependencies [fedb55e]
- @ai-sdk/gateway@1.0.0-beta.16
## 5.0.0-beta.29
### Patch Changes
- e7fcc86: feat (ai): introduce dynamic tools
- d92b9a8: fix(ai): add support for MCP protocol version 2025-06-18
- Updated dependencies [e7fcc86]
- @ai-sdk/provider-utils@3.0.0-beta.7
- @ai-sdk/gateway@1.0.0-beta.15
## 5.0.0-beta.28
### Patch Changes
- 84343eb: fix (ui): call sendAutomaticallyWhen with updated messages
- a753b3a: feat (provider/anthropic): cache control for tools
- Updated dependencies [ac34802]
- @ai-sdk/provider-utils@3.0.0-beta.6
- @ai-sdk/gateway@1.0.0-beta.14
## 5.0.0-beta.27
### Patch Changes
- d5ae088: feat (ui): add sendAutomaticallyWhen to Chat
- Updated dependencies [0477a13]
- Updated dependencies [cf1e00e]
- Updated dependencies [cc21603]
- @ai-sdk/gateway@1.0.0-beta.13
## 5.0.0-beta.26
### Patch Changes
- ae77a99: chore (ai): rename text and reasoning chunks in streamText fullstream
- 1f6ce57: feat (ai): infer tool call types in the `onToolCall` callback
- 5a975a4: feat (ui): update Chat tool result submission
- 2a62513: Fix error thrown when emptying messages in onError or onFinish
- 904fa5e: feat (ai/core): add terminateOnError option to readUIMessageStream
- f81c720: chore(ai): bundle dependencies in CLI binary
- Updated dependencies [70ebead]
- @ai-sdk/gateway@1.0.0-beta.12
## 5.0.0-beta.25
### Patch Changes
- Updated dependencies [8bd3624]
- Updated dependencies [e001ea1]
- @ai-sdk/gateway@1.0.0-beta.11
## 5.0.0-beta.24
### Patch Changes
- add5ac1: feat (ai): make streamText toUIMessageStream async iterable
- ff1c81a: feat (ai): add streamText onAbort callback
- e4c8647: feat (ui): allow asynchronous onFinish in createUIMessageStream
- 383cbfa: feat (ai): add isAborted to onFinish callback for ui message streams
- Updated dependencies [57edfcb]
- Updated dependencies [383cbfa]
- @ai-sdk/provider-utils@3.0.0-beta.5
- @ai-sdk/gateway@1.0.0-beta.10
## 5.0.0-beta.23
### Patch Changes
- 89ba235: fix (ui): support tool names with dash
## 5.0.0-beta.22
### Patch Changes
- de2d2ab: feat(ai): add provider and provider registry middleware functionality
- 6c42e56: feat (ai): validate ui stream data chunks
- c93a8bc: chore(ai): export AsyncIterableStream type from async-iterable-stream module
- 20398f2: feat: ai sdk cli documentation + adjusted default model
- 86293e5: fix (ai): use correct generateMessageId in streamText toUIMessageStream
- 205077b: fix: improve Zod compatibility
- Updated dependencies [205077b]
- @ai-sdk/provider-utils@3.0.0-beta.4
- @ai-sdk/gateway@1.0.0-beta.9
## 5.0.0-beta.21
### Patch Changes
- 38ae5cc: feat (ai): export InferUIMessageChunk type
- faea29f: fix (provider/openai): multi-step reasoning with text
- 90ac328: fix (ui): tool part metadata support in ui messages
- 4a1e0c8: fix(ai-cli): fix bundling and improve authentication error handling
- 30ac566: fix (ui): text message metadata support in ui messages
## 5.0.0-beta.20
### Patch Changes
- 4c8f834: feat: automatically respect rate limit headers in retry logic
Added automatic support for respecting rate limit headers (`retry-after-ms` and `retry-after`) in the SDK's retry logic. When these headers are present and contain reasonable values (0-60 seconds), the retry mechanism will use the server-specified delay instead of exponential backoff. This matches the behavior of Anthropic and OpenAI client SDKs and improves rate limit handling without requiring any API changes.
## 5.0.0-beta.19
### Patch Changes
- 10b21eb: feat(cli): add ai command line interface
- 75c3396: fix (ai): handle errors in 2nd streamText doStream call
- 05d2819: feat: allow zod 4.x as peer dependency
- db64cbe: fix (provider/openai): multi-step reasoning with tool calls
- Updated dependencies [05d2819]
- @ai-sdk/provider-utils@3.0.0-beta.3
- @ai-sdk/gateway@1.0.0-beta.8
## 5.0.0-beta.18
### Patch Changes
- d3960e3: selectTelemetryAttributes more robustness
- 9338f3e: fix (ai): throw error for v1 models
## 5.0.0-beta.17
### Patch Changes
- Updated dependencies [c190907]
- @ai-sdk/gateway@1.0.0-beta.7
## 5.0.0-beta.16
### Patch Changes
- Updated dependencies [9e16bfd]
- @ai-sdk/gateway@1.0.0-beta.6
## 5.0.0-beta.15
### Patch Changes
- 8e31d46: feat (ai): export SourceDocumentUIPart
## 5.0.0-beta.14
### Patch Changes
- Updated dependencies [30ab1de]
- @ai-sdk/gateway@1.0.0-beta.5
## 5.0.0-beta.13
### Patch Changes
- 377bbcf: fix (ui): tool input can be undefined during input-streaming
- ce1d1f3: feat (ai): export mock image, speech, and transcription models
- c040e2f: fix (ui): inject generated response message id
- c808e4d: fix (ui): do not send changing assistant message ids when onFinish is provided
## 5.0.0-beta.12
### Patch Changes
- fc0380b: feat (ui): resolvable header, body, credentials in http chat transport
- 51f497d: feat (ai): step input message modification in prepareStep
- 4f3776c: feat (ai): add InferUITools helper
## 5.0.0-beta.11
### Patch Changes
- 9e40cbe: Allow destructuring output and errorText on `ToolUIPart` type
## 5.0.0-beta.10
### Major Changes
- 2b637d6: chore (ai): rename UIMessageStreamPart to UIMessageChunk
### Patch Changes
- 16ccfb2: feat (ai): add readUIMessageStream helper
- 90ca2b9: feat(ai): Record tool call errors on tool call spans recorded in `generateText` and `streamText`.
- af1d5a5: fix(ai): Unexpected reasoning-start event in extract reasoning middleware
## 5.0.0-beta.9
### Patch Changes
- 86cfc72: feat (ai): add ignoreIncompleteToolCalls option to convertToModelMessages
## 5.0.0-beta.8
### Patch Changes
- 6909543: feat (ai): support system parameter in Agent constructor
- c8fce91: feat (ai): add experimental Agent abstraction
- 9121250: Expose provider metadata as an attribute on exported OTEL spans
- Updated dependencies [97fedf9]
- @ai-sdk/gateway@1.0.0-beta.4
## 5.0.0-beta.7
### Patch Changes
- 60132dd: fixed date formatting for updated mcp protocol version
## 5.0.0-beta.6
### Patch Changes
- 143c55b: feat (ai): export Chat callback types
- f04ffe4: feat (ui): add onData callback to Chat
- 97c35c0: feat (ui): transient data parts
- fccf75c: update mcp protocol version
## 5.0.0-beta.5
### Patch Changes
- 4f3e637: fix (ui): avoid caching globalThis.fetch in case it is patched by other libraries
## 5.0.0-beta.4
### Patch Changes
- 09f41ac: fix (ui): add message metadata in Chat.sendMessage
## 5.0.0-beta.3
### Patch Changes
- Updated dependencies [f3639fa]
- Updated dependencies [d454e4b]
- @ai-sdk/gateway@1.0.0-beta.3
## 5.0.0-beta.2
### Patch Changes
- 0d9583c: fix (ai): use user-provided media type when available
- c6b64a7: feat (ai): allow async prepareRequest on HttpChatTransport
- cb3b9c9: fix (ai): catch errors in ui message stream
- d1a034f: feature: using Zod 4 for internal stuff
- Updated dependencies [0571b98]
- Updated dependencies [c91586a]
- Updated dependencies [39a4fab]
- Updated dependencies [d1a034f]
- @ai-sdk/provider-utils@3.0.0-beta.2
- @ai-sdk/gateway@1.0.0-beta.2
## 5.0.0-beta.1
### Major Changes
- 9ad0484: feat (ai): automatic tool execution error handling
### Patch Changes
- d88455d: feat (ai): expose http chat transport type
- 4048ce3: fix (ai): add tests and examples for openai responses
- f2b041e: Fix custom `fetch` in HttpChatTransport
- cb68df0: feat: add transcription and speech model support to provider registry
- 26695a3: feat (ui): add state for text and reasoning ui message parts
- e7d2ce3: feat: provider-executed tools
- 102b066: fix (ai): fix invalid fetch call
- e862b5b: feat (ai): allow sync tool.execute
- 7bd025b: fix (ai): fix sync tool execute with streamText
- Updated dependencies [742b7be]
- Updated dependencies [7cddb72]
- Updated dependencies [ccce59b]
- Updated dependencies [e2b9e4b]
- Updated dependencies [45c1ea2]
- Updated dependencies [e025824]
- Updated dependencies [0d06df6]
- Updated dependencies [472524a]
- Updated dependencies [dd3ff01]
- Updated dependencies [7435eb5]
- Updated dependencies [cb68df0]
- Updated dependencies [bfdca8d]
- Updated dependencies [f77bc38]
- Updated dependencies [44f4aba]
- Updated dependencies [023ba40]
- Updated dependencies [5e57fae]
- Updated dependencies [71f938d]
- Updated dependencies [28a5ed5]
- @ai-sdk/provider@2.0.0-beta.1
- @ai-sdk/provider-utils@3.0.0-beta.1
- @ai-sdk/gateway@1.0.0-beta.1
## 5.0.0-alpha.15
### Major Changes
- 04d5063: chore (ai): rename default provider global to AI_SDK_DEFAULT_PROVIDER
- b4b4bb2: chore (ui): rename experimental_resume to resumeStream
- d884051: feat (ai): simplify default provider setup
- 954aa73: feat (ui): extended regenerate support
- 60e2c56: feat (ai): restructure chat transports
### Patch Changes
- b1e3abd: feat (ai): expose ui message stream headers
- 142576e: feat (ui): support message replacement in chat via messageId param on sendMessage
- 395c85e: feat (ai): add consumeSseStream option to UI message stream responses
- Updated dependencies [48d257a]
- Updated dependencies [8ba77a7]
- Updated dependencies [c145d62]
- @ai-sdk/provider@2.0.0-alpha.15
- @ai-sdk/provider-utils@3.0.0-alpha.15
- @ai-sdk/gateway@1.0.0-alpha.15
## 5.0.0-alpha.14
### Major Changes
- 63f9e9b: chore (provider,ai): tools have input/output instead of args,result
### Patch Changes
- Updated dependencies [b5da06a]
- Updated dependencies [63f9e9b]
- Updated dependencies [2e13791]
- @ai-sdk/provider@2.0.0-alpha.14
- @ai-sdk/gateway@1.0.0-alpha.14
- @ai-sdk/provider-utils@3.0.0-alpha.14
## 5.0.0-alpha.13
### Major Changes
- 0a710d8: feat (ui): typed tool parts in ui messages
- 6a83f7d: refactoring (ai): restructure message metadata transfer
- 1f55c21: chore (ai): send reasoning to the client by default
- 33eb499: feat (ai): inject message id in createUIMessageStream
- 901df02: feat (ui): use UI_MESSAGE generic
### Patch Changes
- Updated dependencies [68ecf2f]
- @ai-sdk/provider@2.0.0-alpha.13
- @ai-sdk/gateway@1.0.0-alpha.13
- @ai-sdk/provider-utils@3.0.0-alpha.13
## 5.0.0-alpha.12
### Major Changes
- 4892798: chore (ai): always stream tool calls
### Patch Changes
- da1e6f0: feat (ui): add generics to ui message stream parts
- Updated dependencies [e2aceaf]
- @ai-sdk/gateway@1.0.0-alpha.12
- @ai-sdk/provider@2.0.0-alpha.12
- @ai-sdk/provider-utils@3.0.0-alpha.12
## 5.0.0-alpha.11
### Major Changes
- e8324c5: feat (ai): add args callbacks to tools
### Patch Changes
- Updated dependencies [c1e6647]
- @ai-sdk/provider@2.0.0-alpha.11
- @ai-sdk/gateway@1.0.0-alpha.11
- @ai-sdk/provider-utils@3.0.0-alpha.11
## 5.0.0-alpha.10
### Major Changes
- 98f25e5: chore (ui): remove managed chat inputs
- 7bb58d4: chore (ai): restructure prepareRequest
### Patch Changes
- Updated dependencies [c4df419]
- @ai-sdk/provider@2.0.0-alpha.10
- @ai-sdk/gateway@1.0.0-alpha.10
- @ai-sdk/provider-utils@3.0.0-alpha.10
## 5.0.0-alpha.9
### Major Changes
- 9ae327d: chore (ui): replace chat store concept with chat instances
### Patch Changes
- 8255639: ### Fix use with Google APIs + zod v4's `.literal()` schema
Before [zod@3.25.49](https://github.com/colinhacks/zod/releases/tag/v3.25.49), requests to Google's APIs failed due to a missing `type` in the provided schema. The problem has been resolved for the `ai` SDK by bumping our `zod` peer dependencies to `^3.25.49`.
pull request: https://github.com/vercel/ai/pull/6609
- Updated dependencies [26b6dd0]
- Updated dependencies [811dff3]
- @ai-sdk/gateway@1.0.0-alpha.9
- @ai-sdk/provider@2.0.0-alpha.9
- @ai-sdk/provider-utils@3.0.0-alpha.9
## 5.0.0-alpha.8
### Major Changes
- c25cbce: feat (ai): use console.error as default error handler for streamText and streamObject
### Patch Changes
- 4fef487: feat: support for zod v4 for schema validation
All these methods now accept both a zod v4 and zod v3 schemas for validation:
- `generateObject()`
- `streamObject()`
- `generateText()`
- `experimental_useObject()` from `@ai-sdk/react`
- `streamUI()` from `@ai-sdk/rsc`
- 6b1c55c: feat (ai): introduce GLOBAL_DEFAULT_PROVIDER
- 2e4f9e4: feat (ai): improved error messages when using gateway
- Updated dependencies [4fef487]
- Updated dependencies [9222aeb]
- Updated dependencies [3cbcbb7]
- Updated dependencies [989ac75]
- Updated dependencies [7742ba3]
- @ai-sdk/provider-utils@3.0.0-alpha.8
- @ai-sdk/provider@2.0.0-alpha.8
- @ai-sdk/gateway@1.0.0-alpha.8
## 5.0.0-alpha.7
### Major Changes
- db345da: chore (ai): remove exports of internal ui functions
- 247ee0c: chore (ai): remove steps from tool invocation ui parts
### Patch Changes
- 9b0da33: fix (ai): do not send id with start unless specified
- Updated dependencies [5c56081]
- @ai-sdk/provider@2.0.0-alpha.7
- @ai-sdk/gateway@1.0.0-alpha.7
- @ai-sdk/provider-utils@3.0.0-alpha.7
## 5.0.0-alpha.6
### Patch Changes
- 0d2c085: feat (ai): support string model ids through gateway
- 48a7606: feat (ai): support changing the system prompt in prepareSteps
- Updated dependencies [0d2c085]
- Updated dependencies [6c2c708]
- @ai-sdk/provider@2.0.0-alpha.6
- @ai-sdk/gateway@1.0.0-alpha.6
- @ai-sdk/provider-utils@3.0.0-alpha.6
## 5.0.0-alpha.5
### Major Changes
- ef256ed: chore (ai): refactor and use chatstore in svelte
- 1ed0287: chore (ai): stable sendStart/sendFinish options
### Patch Changes
- 655cf3c: feat (ui): add onFinish to createUIMessageStream
- 1675396: fix: avoid job executor deadlock when adding tool result
- cf9af6e: feat (ai): allow sync prepareStep
- 825e8d7: release alpha.5
- 7324c21: fix (ai/telemetry): Avoid JSON.stringify on Uint8Arrays for telemetry
## 5.0.0-alpha.4
### Major Changes
- 72d7d72: chore (ai): stable activeTools
- 9315076: chore (ai): rename continueUntil to stopWhen. Rename maxSteps stop condition to stepCountIs.
### Patch Changes
- b32c141: feat (ai): add array support to stopWhen
- 7d97ab6: release alpha.4
- 37a916d: feat (ai): add prepareSteps to streamText
- 5f2b3d4: chore (ai): stable prepareStep
- Updated dependencies [dc714f3]
- @ai-sdk/provider@2.0.0-alpha.4
- @ai-sdk/provider-utils@3.0.0-alpha.4
## 5.0.0-alpha.3
### Major Changes
- ab7ccef: chore (ai): change source ui message parts to source-url
- 257224b: chore (ai): separate TextStreamChatTransport
- 0463011: fix (ai): update source url stream part
- d306260: feat (ai): replace maxSteps with continueUntil (streamText)
### Patch Changes
- Updated dependencies [6b98118]
- @ai-sdk/provider@2.0.0-alpha.3
- @ai-sdk/provider-utils@3.0.0-alpha.3
## 5.0.0-alpha.2
### Patch Changes
- 82aa95d: fix (ai): merge data ui stream parts correctly
- Updated dependencies [26535e0]
- @ai-sdk/provider@2.0.0-alpha.2
- @ai-sdk/provider-utils@3.0.0-alpha.2
## 5.0.0-alpha.1
### Major Changes
- 109c0ac: chore (ai): rename id to chatId (in post request, resume request, and useChat)
### Patch Changes
- b346545: feat (ai): add data ui part schemas
- Updated dependencies [3f2f00c]
- @ai-sdk/provider@2.0.0-alpha.1
- @ai-sdk/provider-utils@3.0.0-alpha.1
## 5.0.0-canary.24
### Major Changes
- f7e8bf4: chore (ai): flatten ui message stream parts
- ed675de: feat (ai): add ui data parts
- 64f6d64: feat (ai): replace maxSteps with continueUntil (generateText)
### Patch Changes
- bedb239: chore (ai): make ui stream parts value optional when it's not required
- 507ac1d: fix (ui/react): update messages immediately with the submitted user message
- 2b9bbcd: feat (ai): improve prompt validation error message
- cda32ba: fix (ai): send `start` part in correct position in stream (streamText)
- 50f0362: fix (ai): fix experimental sendStart/sendFinish options in streamText
- Updated dependencies [faf8446]
- @ai-sdk/provider-utils@3.0.0-canary.19
## 5.0.0-canary.23
### Major Changes
- 40acf9b: feat (ui): introduce ChatStore and ChatTransport
### Patch Changes
- Updated dependencies [40acf9b]
- @ai-sdk/provider-utils@3.0.0-canary.18
## 5.0.0-canary.22
### Major Changes
- e7dc6c7: chore (ai): remove onResponse callback
- a34eb39: chore (ai): remove `data` and `allowEmptySubmit` from `ChatRequestOptions`
- b33ed7a: chore (ai): rename DataStream* to UIMessage*
- 765f1cd: chore (ai): remove deprecated useChat isLoading helper
## 5.0.0-canary.21
### Major Changes
- d964901: - remove setting temperature to `0` by default
- remove `null` option from `DefaultSettingsMiddleware`
- remove setting defaults for `temperature` and `stopSequences` in `ai` to enable middleware changes
- 0560977: chore (ai): improve consistency of generate text result, stream text result, and step result
- 516be5b: ### Move Image Model Settings into generate options
Image Models no longer have settings. Instead, `maxImagesPerCall` can be passed directly to `generateImage()`. All other image settings can be passed to `providerOptions[provider]`.
Before
```js
await generateImage({
model: luma.image('photon-flash-1', {
maxImagesPerCall: 5,
pollIntervalMillis: 500,
}),
prompt,
n: 10,
});
```
After
```js
await generateImage({
model: luma.image('photon-flash-1'),
prompt,
n: 10,
maxImagesPerCall: 5,
providerOptions: {
luma: { pollIntervalMillis: 5 },
},
});
```
Pull Request: https://github.com/vercel/ai/pull/6180
- bfbfc4c: feat (ai): streamText/generateText: totalUsage contains usage for all steps. usage is for a single step.
- ea7a7c9: feat (ui): UI message metadata
- 1409e13: chore (ai): remove experimental continueSteps
### Patch Changes
- 66af894: fix (ai): respect content order in toResponseMessages
- Updated dependencies [ea7a7c9]
- @ai-sdk/provider-utils@3.0.0-canary.17
## 5.0.0-canary.20
### Major Changes
- 13fef90: chore (ai): remove automatic conversion of UI messages to model messages
- 496bbc1: chore (ui): inline/remove ChatRequest type
- da70d79: chore (ai): remove getUIText helper
- c7710a9: chore (ai): rename DataStreamToSSETransformStream to JsonToSseTransformStream
- 35fc02c: chore (ui): rename RequestOptions to CompletionRequestOptions
- b983b51: feat (ai): support model message array in prompt
### Minor Changes
- bcea599: feat (ai): add content to generateText result
- 48d675a: feat (ai): add content to streamText result
### Patch Changes
- e90d45d: chore (rsc): move HANGING_STREAM_WARNING_TIME constant into @ai-sdk/rsc package
- bc3109f: chore (ai): push stream-callbacks into langchain/llamaindex adapters
- Updated dependencies [87b828f]
- @ai-sdk/provider-utils@3.0.0-canary.16
## 5.0.0-canary.19
### Major Changes
- 2d03e19: chore (ai): remove StreamCallbacks.onCompletion
- 319b989: chore (ai): remove content from ui messages
- 441d042: chore (ui): data stream protocol v2 with SSEs
- dcc549b: remove StreamTextResult.mergeIntoDataStream method
rename DataStreamOptions.getErrorMessage to onError
add pipeTextStreamToResponse function
add createTextStreamResponse function
change createDataStreamResponse function to accept a DataStream and not a DataStreamWriter
change pipeDataStreamToResponse function to accept a DataStream and not a DataStreamWriter
change pipeDataStreamToResponse function to have a single parameter
- cb2b53a: chore (ai): refactor header preparation
- e244a78: chore (ai): remove StreamData and mergeStreams
## 5.0.0-canary.18
### Major Changes
- c60f895: chore (ai): remove useChat keepLastMessageOnError
- a662dea: chore (ai): remove sendExtraMessageFields
### Patch Changes
- a571d6e: chore(provider-utils): move ToolResultContent to provider-utils
- 332167b: chore (ai): move maxSteps into UseChatOptions
- a8c8bd5: feat(embed-many): respect supportsParallelCalls & concurrency
- Updated dependencies [a571d6e]
- Updated dependencies [a8c8bd5]
- Updated dependencies [7979f7f]
- Updated dependencies [41fa418]
- @ai-sdk/provider-utils@3.0.0-canary.15
- @ai-sdk/provider@2.0.0-canary.14
## 5.0.0-canary.17
### Major Changes
- f04fb4a: chore (ai): replace useChat attachments with file ui parts
- fd1924b: chore (ai): remove redundant `mimeType` property
- fafc3f2: chore (ai): change file to parts to use urls instead of data
- 92cb0a2: chore (ai): rename CoreMessage to ModelMessage
### Minor Changes
- c9ad635: feat (ai): add filename to file ui parts
### Patch Changes
- 9bd5ab5: feat (provider): add providerMetadata to ImageModelV2 interface (#5977)
The `experimental_generateImage` method from the `ai` package now returnes revised prompts for OpenAI's image models.
```js
const prompt = 'Santa Claus driving a Cadillac';
const { providerMetadata } = await experimental_generateImage({
model: openai.image('dall-e-3'),
prompt,
});
const revisedPrompt = providerMetadata.openai.images[0]?.revisedPrompt;
console.log({
prompt,
revisedPrompt,
});
```
- Updated dependencies [957b739]
- Updated dependencies [9bd5ab5]
- @ai-sdk/provider-utils@3.0.0-canary.14
- @ai-sdk/provider@2.0.0-canary.13
## 5.0.0-canary.16
### Major Changes
- ec78cdc: chore (ai): remove "data" UIMessage role
- 8b86e99: chore (ai): replace `Message` with `UIMessage`
- 2524fc7: chore (ai): remove ui message toolInvocations property
- 175b868: chore (ai): rename reasoning UI parts 'reasoning' property to 'text'
### Patch Changes
- 9b4d074: feat(streamObject): add enum support
- 28ad69e: fix(react-native): support experimental_attachments without FileList global
- ec5933d: chore (ai/mcp): add `assertCapability` method to experimental MCP client
## 5.0.0-canary.15
### Major Changes
- 4bfe9ec: chore (ai): remove ui message reasoning property
- 2877a74: chore (ai): remove ui message data property
### Patch Changes
- d9209ca: fix (image-model): `specificationVersion: v1` -> `v2`
- ea27cc6: chore (ai): use JSONValue definition from provider
- 0ff02bb: chore(provider-utils): move over jsonSchema
- Updated dependencies [7b3ae3f]
- Updated dependencies [0ff02bb]
- @ai-sdk/provider@2.0.0-canary.12
- @ai-sdk/provider-utils@3.0.0-canary.13
## 5.0.0-canary.14
### Patch Changes
- 9bf7291: chore(providers/openai): enable structuredOutputs by default & switch to provider option
- 4617fab: chore(embedding-models): remove remaining settings
- a76a62b: feat (ai): add experimental prepareStep callback to generateText
- Updated dependencies [9bf7291]
- Updated dependencies [4617fab]
- Updated dependencies [e030615]
- @ai-sdk/provider@2.0.0-canary.11
- @ai-sdk/provider-utils@3.0.0-canary.12
## 5.0.0-canary.13
### Patch Changes
- 14cb3be: chore(providers/llamaindex): extract to separate package
- 66962ed: fix(packages): export node10 compatible types
- 9301f86: refactor (image-model): rename `ImageModelV1` to `ImageModelV2`
- Updated dependencies [66962ed]
- Updated dependencies [9301f86]
- Updated dependencies [a3f768e]
- @ai-sdk/provider-utils@3.0.0-canary.11
- @ai-sdk/provider@2.0.0-canary.10
## 5.0.0-canary.12
### Patch Changes
- Updated dependencies [e86be6f]
- @ai-sdk/provider@2.0.0-canary.9
- @ai-sdk/provider-utils@3.0.0-canary.10
## 5.0.0-canary.11
### Patch Changes
- 8e64e9c: feat (ai): allow using provider default temperature by specifying null
- Updated dependencies [95857aa]
- Updated dependencies [7ea4132]
- @ai-sdk/provider@2.0.0-canary.8
- @ai-sdk/provider-utils@3.0.0-canary.9
## 5.0.0-canary.10
### Patch Changes
- d8aeaef: feat(providers/fal): add transcribe
- 3e10408: fix(utils/detect-mimetype): add support for detecting id3 tags
## 5.0.0-canary.9
### Major Changes
- a847c3e: chore: rename reasoning to reasoningText etc
- b32e192: chore (ai): rename reasoning to reasoningText, rename reasoningDetails to reasoning (streamText, generateText)
### Patch Changes
- cb9c9e4: remove deprecated `experimental_wrapLanguageModel`
- 8aa9e20: feat: add speech with experimental_generateSpeech
- Updated dependencies [5d142ab]
- Updated dependencies [b6b43c7]
- Updated dependencies [8aa9e20]
- Updated dependencies [3795467]
- @ai-sdk/provider-utils@3.0.0-canary.8
- @ai-sdk/provider@2.0.0-canary.7
## 5.0.0-canary.8
### Major Changes
- 14c9410: chore: refactor file towards source pattern (spec)
### Patch Changes
- 5d1e3ba: chore (ai): remove provider re-exports
- 26735b5: chore(embedding-model): add v2 interface
- 7827a49: fix (ai/core): refactor `toResponseMessages` to filter out empty string/content
- bd8a36c: feat(embedding-model-v2/embedMany): add response body field
- b6f9f3c: remove deprecated `CoreTool*` types
- 92c8e66: fix(ai/core): properly handle custom separator in provider registry
- fd65bc6: chore(embedding-model-v2): rename rawResponse to response
- 5bdff05: Removed deprecated `options.throwErrorForEmptyVectors` from `cosineSimilarity()`. Since `throwErrorForEmptyVectors` was the only option the entire `options` argument was removed.
```diff
- cosineSimilarity(vector1, vector2, options)
+cosineSimilarity(vector1, vector2)
```
- Updated dependencies [26735b5]
- Updated dependencies [443d8ec]
- Updated dependencies [14c9410]
- Updated dependencies [d9c98f4]
- Updated dependencies [c4a2fec]
- Updated dependencies [0054544]
- Updated dependencies [9e9c809]
- Updated dependencies [32831c6]
- Updated dependencies [d0f9495]
- Updated dependencies [fd65bc6]
- Updated dependencies [393138b]
- Updated dependencies [7182d14]
- @ai-sdk/provider@2.0.0-canary.6
- @ai-sdk/provider-utils@3.0.0-canary.7
## 5.0.0-canary.7
### Major Changes
- 6fba4c7: chore (ai): remove deprecated experimental_providerMetadata
- 1766ede: chore: rename maxTokens to maxOutputTokens
### Patch Changes
- 0b78e17: chore(ai/generateObject): simplify function signature
- 3e3b9df: fix (ai/mcp): better support for zero-argument MCP tools
- f10304b: feat(tool-calling): don't require the user to have to pass parameters
- Updated dependencies [411e483]
- Updated dependencies [79457bd]
- Updated dependencies [ad80501]
- Updated dependencies [1766ede]
- Updated dependencies [f10304b]
- @ai-sdk/provider@2.0.0-canary.5
- @ai-sdk/provider-utils@3.0.0-canary.6
## 5.0.0-canary.6
### Patch Changes
- Updated dependencies [6f6bb89]
- @ai-sdk/provider@2.0.0-canary.4
- @ai-sdk/provider-utils@3.0.0-canary.5
## 5.0.0-canary.5
### Patch Changes
- b71fe8d: fix(ai): remove jsondiffpatch dependency
- d91b50d: chore(ui-utils): merge into ai package
- Updated dependencies [d1a1aa1]
- @ai-sdk/provider@2.0.0-canary.3
- @ai-sdk/provider-utils@3.0.0-canary.4
## 5.0.0-canary.4
### Major Changes
- e1cbf8a: chore(@ai-sdk/rsc): extract to separate package
### Patch Changes
- 225f087: fix (ai/mcp): prevent mutation of customEnv
- a166433: feat: add transcription with experimental_transcribe
- 0a87932: core (ai): change transcription model mimeType to mediaType
- Updated dependencies [a166433]
- Updated dependencies [abf9a79]
- Updated dependencies [9f95b35]
- Updated dependencies [0a87932]
- Updated dependencies [6dc848c]
- @ai-sdk/provider-utils@3.0.0-canary.3
- @ai-sdk/provider@2.0.0-canary.2
- @ai-sdk/ui-utils@2.0.0-canary.3
## 5.0.0-canary.3
### Patch Changes
- Updated dependencies [c57e248]
- Updated dependencies [33f4a6a]
- @ai-sdk/provider@2.0.0-canary.1
- @ai-sdk/provider-utils@3.0.0-canary.2
- @ai-sdk/ui-utils@2.0.0-canary.2
## 5.0.0-canary.2
### Patch Changes
- bd398e4: fix (core): improve error handling in streamText's consumeStream method
## 5.0.0-canary.1
### Minor Changes
- b7eae2d: feat (core): Add finishReason field to NoObjectGeneratedError
### Patch Changes
- c22ad54: feat(smooth-stream): chunking callbacks
- a4f3007: chore: remove ai/react
- Updated dependencies [060370c]
- Updated dependencies [0c0c0b3]
- Updated dependencies [63d791d]
- @ai-sdk/provider-utils@3.0.0-canary.1
- @ai-sdk/ui-utils@2.0.0-canary.1
## 5.0.0-canary.0
### Major Changes
- d5f588f: AI SDK 5
- 9477ebb: chore (ui): remove useAssistant hook (**breaking change**)
### Patch Changes
- 8026705: fix (core): send buffered text in smooth stream when stream parts change
- Updated dependencies [d5f588f]
- Updated dependencies [9477ebb]
- @ai-sdk/provider-utils@3.0.0-canary.0
- @ai-sdk/ui-utils@2.0.0-canary.0
- @ai-sdk/react@2.0.0-canary.0
- @ai-sdk/provider@2.0.0-canary.0
## 4.2.10
### Patch Changes
- Updated dependencies [a043b14]
- Updated dependencies [28be004]
- @ai-sdk/react@1.2.5
- @ai-sdk/provider-utils@2.2.3
- @ai-sdk/ui-utils@1.2.4
## 4.2.9
### Patch Changes
- Updated dependencies [b01120e]
- @ai-sdk/provider-utils@2.2.2
- @ai-sdk/react@1.2.4
- @ai-sdk/ui-utils@1.2.3
## 4.2.8
### Patch Changes
- 65243ce: fix (ui): introduce step start parts
- Updated dependencies [65243ce]
- @ai-sdk/ui-utils@1.2.2
- @ai-sdk/react@1.2.3
## 4.2.7
### Patch Changes
- e14c066: fix (ai/core): convert user ui messages with only parts (no content) to core messages
## 4.2.6
### Patch Changes
- 625591b: feat (ai/core): auto-complete for provider registry
- 6a1506f: feat (ai/core): custom separator support for provider registry
- ea3d998: chore (ai/core): move provider registry to stable
## 4.2.5
### Patch Changes
- Updated dependencies [d92fa29]
- @ai-sdk/react@1.2.2
## 4.2.4
### Patch Changes
- 3d6d96d: fix (ai/core): validate that messages are not empty
## 4.2.3
### Patch Changes
- 0b3bf29: fix (ai/core): custom env support for stdio MCP transport
## 4.2.2
### Patch Changes
- f10f0fa: fix (provider-utils): improve event source stream parsing performance
- Updated dependencies [f10f0fa]
- @ai-sdk/provider-utils@2.2.1
- @ai-sdk/react@1.2.1
- @ai-sdk/ui-utils@1.2.1
## 4.2.1
### Patch Changes
- b796152: feat (ai/core): add headers to MCP SSE transport
- 06361d6: feat (ai/core): expose JSON RPC types (MCP)
## 4.2.0
### Minor Changes
- 5bc638d: AI SDK 4.2
### Patch Changes
- Updated dependencies [5bc638d]
- @ai-sdk/provider@1.1.0
- @ai-sdk/provider-utils@2.2.0
- @ai-sdk/react@1.2.0
- @ai-sdk/ui-utils@1.2.0
## 4.1.66
### Patch Changes
- 5d0fc29: chore (ai): improve cosine similarity calculation
## 4.1.65
### Patch Changes
- 16c444f: fix (ai): expose ai/mcp-stdio
## 4.1.64
### Patch Changes
- Updated dependencies [d0c4659]
- @ai-sdk/provider-utils@2.1.15
- @ai-sdk/react@1.1.25
- @ai-sdk/ui-utils@1.1.21
## 4.1.63
### Patch Changes
- 0bd5bc6: feat (ai): support model-generated files
- Updated dependencies [0bd5bc6]
- @ai-sdk/provider@1.0.12
- @ai-sdk/provider-utils@2.1.14
- @ai-sdk/ui-utils@1.1.20
- @ai-sdk/react@1.1.24
## 4.1.62
### Patch Changes
- c9ed3c4: feat: enable custom mcp transports
breaking change: remove internal stdio transport creation
## 4.1.61
### Patch Changes
- 2e1101a: feat (provider/openai): pdf input support
- Updated dependencies [2e1101a]
- @ai-sdk/provider@1.0.11
- @ai-sdk/provider-utils@2.1.13
- @ai-sdk/ui-utils@1.1.19
- @ai-sdk/react@1.1.23
## 4.1.60
### Patch Changes
- 0b8797f: feat (ai/core): expose response body for each generateText step
## 4.1.59
### Patch Changes
- dd18049: fix (ai/core): suppress next.js warnings for node.js specific code path
## 4.1.58
### Patch Changes
- e9897eb: fix (ai/core): move process access into functions and use globalThis
## 4.1.57
### Patch Changes
- 092fdaa: feat (ai/core): add defaultSettingsMiddleware
## 4.1.56
### Patch Changes
- 80be82b: feat (ai/core): add simulateStreamingMiddleware
- 8109a24: fix (ai/core): limit node imports to types where possible
## 4.1.55
### Patch Changes
- 1531959: feat (ai/core): add MCP client for using MCP tools
- Updated dependencies [1531959]
- @ai-sdk/provider-utils@2.1.12
- @ai-sdk/react@1.1.22
- @ai-sdk/ui-utils@1.1.18
## 4.1.54
### Patch Changes
- ee1c787: fix (ai/core): correct spread apply order to fix extract reasoning middleware with generateText
## 4.1.53
### Patch Changes
- e1d3d42: feat (ai): expose raw response body in generateText and generateObject
- Updated dependencies [e1d3d42]
- @ai-sdk/provider@1.0.10
- @ai-sdk/provider-utils@2.1.11
- @ai-sdk/ui-utils@1.1.17
- @ai-sdk/react@1.1.21
## 4.1.52
### Patch Changes
- 5329a69: fix (ai/core): fix duplicated reasoning in streamText onFinish and messages
## 4.1.51
### Patch Changes
- 0cb2647: feat (ai/core): add streamText sendStart & sendFinish data stream options
## 4.1.50
### Patch Changes
- ae98f0d: fix (ai/core): forward providerOptions for text, image, and file parts
## 4.1.49
### Patch Changes
- dc027d3: fix (ai/core): add reasoning support to appendResponseMessages
## 4.1.48
### Patch Changes
- Updated dependencies [6255fbc]
- @ai-sdk/react@1.1.20
## 4.1.47
### Patch Changes
- Updated dependencies [da5c734]
- @ai-sdk/react@1.1.19
## 4.1.46
### Patch Changes
- ddf9740: feat (ai): add anthropic reasoning
- Updated dependencies [ddf9740]
- @ai-sdk/provider@1.0.9
- @ai-sdk/ui-utils@1.1.16
- @ai-sdk/provider-utils@2.1.10
- @ai-sdk/react@1.1.18
## 4.1.45
### Patch Changes
- 93bd5a0: feat (ai/ui): add writeSource to createDataStream
## 4.1.44
### Patch Changes
- f8e7df2: fix (ai/core): add `startWithReasoning` option to `extractReasoningMiddleware`
## 4.1.43
### Patch Changes
- ef2e23b: feat (ai/core): add experimental repairText function to generateObject
## 4.1.42
### Patch Changes
- Updated dependencies [2761f06]
- @ai-sdk/provider@1.0.8
- @ai-sdk/provider-utils@2.1.9
- @ai-sdk/ui-utils@1.1.15
- @ai-sdk/react@1.1.17
## 4.1.41
### Patch Changes
- Updated dependencies [60c3220]
- @ai-sdk/react@1.1.16
## 4.1.40
### Patch Changes
- Updated dependencies [c43df41]
- @ai-sdk/react@1.1.15
## 4.1.39
### Patch Changes
- 075a9a9: fix (ai): improve tsdoc on custom provider
## 4.1.38
### Patch Changes
- 4c9c194: chore (ai): add description to provider-defined tools for better accessibility
- 2e898b4: chore (ai): move mockId test helper into provider utils
- Updated dependencies [2e898b4]
- @ai-sdk/provider-utils@2.1.8
- @ai-sdk/react@1.1.14
- @ai-sdk/ui-utils@1.1.14
## 4.1.37
### Patch Changes
- c1e10d1: chore: export UIMessage type
## 4.1.36
### Patch Changes
- Updated dependencies [3ff4ef8]
- @ai-sdk/provider-utils@2.1.7
- @ai-sdk/react@1.1.13
- @ai-sdk/ui-utils@1.1.13
## 4.1.35
### Patch Changes
- 166e09e: feat (ai/ui): forward source parts to useChat
- Updated dependencies [166e09e]
- @ai-sdk/ui-utils@1.1.12
- @ai-sdk/react@1.1.12
## 4.1.34
### Patch Changes
- dc49119: chore: deprecate ai/react
## 4.1.33
### Patch Changes
- 74f0f0e: chore (ai/core): move providerMetadata to stable
## 4.1.32
### Patch Changes
- c128ca5: fix (ai/core): fix streamText onFinish messages with structured output
## 4.1.31
### Patch Changes
- b30b1cc: feat (ai/core): add onError callback to streamObject
## 4.1.30
### Patch Changes
- 4ee5b6f: fix (core): remove invalid providerOptions from streamObject onFinish callback
## 4.1.29
### Patch Changes
- 605de49: feat (ai/core): export callback types
## 4.1.28
### Patch Changes
- 6eb7fc4: feat (ai/core): url source support
## 4.1.27
### Patch Changes
- Updated dependencies [318b351]
- @ai-sdk/ui-utils@1.1.11
- @ai-sdk/react@1.1.11
## 4.1.26
### Patch Changes
- 34983d4: fix (ai/core): bind supportsUrl when creating wrapper
## 4.1.25
### Patch Changes
- 5a21310: fix (ai/core): use ai types on custom provider to prevent ts error
## 4.1.24
### Patch Changes
- 38142b8: feat (ai/core): introduce streamText consumeStream
## 4.1.23
### Patch Changes
- b08f7c1: fix (ai/core): suppress errors in textStream
## 4.1.22
### Patch Changes
- 2bec72a: feat (ai/core): add onError callback to streamText
## 4.1.21
### Patch Changes
- d387989: feat (ai/core): re-export zodSchema
## 4.1.20
### Patch Changes
- bcc61d4: feat (ui): introduce message parts for useChat
- Updated dependencies [bcc61d4]
- @ai-sdk/ui-utils@1.1.10
- @ai-sdk/react@1.1.10
## 4.1.19
### Patch Changes
- Updated dependencies [6b8cc14]
- @ai-sdk/ui-utils@1.1.9
- @ai-sdk/react@1.1.9
## 4.1.18
### Patch Changes
- 6a1acfe: fix (ai/core): revert '@internal' tag on function definitions due to build impacts
## 4.1.17
### Patch Changes
- 5af8cdb: fix (ai/core): support this reference in model.supportsUrl implementations
## 4.1.16
### Patch Changes
- 7e299a4: feat (ai/core): wrapLanguageModel can apply multiple middlewares
## 4.1.15
### Patch Changes
- d89c3b9: feat (provider): add image model support to provider specification
- d89c3b9: feat (core): type ahead for model ids with custom provider
- 08f54fc: chore (ai/core): move custom provider to stable
- Updated dependencies [d89c3b9]
- @ai-sdk/provider@1.0.7
- @ai-sdk/provider-utils@2.1.6
- @ai-sdk/ui-utils@1.1.8
- @ai-sdk/react@1.1.8
## 4.1.14
### Patch Changes
- ca89615: fix (ai/core): only append assistant response at the end when there is a final user message
## 4.1.13
### Patch Changes
- 999085e: feat (ai/core): add write function to DataStreamWriter
## 4.1.12
### Patch Changes
- 0d2d9bf: fix (ui): single assistant message with multiple tool steps
- Updated dependencies [0d2d9bf]
- Updated dependencies [0d2d9bf]
- @ai-sdk/react@1.1.7
- @ai-sdk/ui-utils@1.1.7
## 4.1.11
### Patch Changes
- 4c58da5: chore (core): move providerOptions to stable
## 4.1.10
### Patch Changes
- bf2c9c6: feat (core): move middleware to stable
## 4.1.9
### Patch Changes
- 3a602ca: chore (core): rename CoreTool to Tool
- Updated dependencies [3a602ca]
- @ai-sdk/provider-utils@2.1.5
- @ai-sdk/ui-utils@1.1.6
- @ai-sdk/react@1.1.6
## 4.1.8
### Patch Changes
- 92f5f36: feat (core): add extractReasoningMiddleware
## 4.1.7
### Patch Changes
- 066206e: feat (provider-utils): move delay to provider-utils from ai
- Updated dependencies [066206e]
- @ai-sdk/provider-utils@2.1.4
- @ai-sdk/react@1.1.5
- @ai-sdk/ui-utils@1.1.5
## 4.1.6
### Patch Changes
- Updated dependencies [39e5c1f]
- @ai-sdk/provider-utils@2.1.3
- @ai-sdk/react@1.1.4
- @ai-sdk/ui-utils@1.1.4
## 4.1.5
### Patch Changes
- 9ce598c: feat (ai/ui): add reasoning support to useChat
- Updated dependencies [9ce598c]
- @ai-sdk/ui-utils@1.1.3
- @ai-sdk/react@1.1.3
## 4.1.4
### Patch Changes
- caaad11: feat (ai/core): re-export languagemodelv1 types for middleware implementations
- caaad11: feat (ai/core): expose TelemetrySettings type
## 4.1.3
### Patch Changes
- 7f30a77: feat (core): export core message schemas
- 4298996: feat (core): add helper for merging single client message
## 4.1.2
### Patch Changes
- 3c5fafa: chore (ai/core): move streamText toolCallStreaming option to stable
- 3a58a2e: feat (ai/core): throw NoImageGeneratedError from generateImage when no predictions are returned.
- Updated dependencies [ed012d2]
- Updated dependencies [6f4d063]
- Updated dependencies [3a58a2e]
- @ai-sdk/provider-utils@2.1.2
- @ai-sdk/react@1.1.2
- @ai-sdk/provider@1.0.6
- @ai-sdk/ui-utils@1.1.2
## 4.1.1
### Patch Changes
- 0a699f1: feat: add reasoning token support
- Updated dependencies [e7a9ec9]
- Updated dependencies [0a699f1]
- @ai-sdk/ui-utils@1.1.1
- @ai-sdk/provider-utils@2.1.1
- @ai-sdk/provider@1.0.5
- @ai-sdk/react@1.1.1
## 4.1.0
### Minor Changes
- 62ba5ad: release: AI SDK 4.1
### Patch Changes
- Updated dependencies [62ba5ad]
- @ai-sdk/provider-utils@2.1.0
- @ai-sdk/react@1.1.0
- @ai-sdk/ui-utils@1.1.0
## 4.0.41
### Patch Changes
- Updated dependencies [44f04d5]
- @ai-sdk/react@1.0.14
## 4.0.40
### Patch Changes
- 33592d2: fix (ai/core): switch to json schema 7 target for zod to json schema conversion
- Updated dependencies [33592d2]
- @ai-sdk/ui-utils@1.0.12
- @ai-sdk/react@1.0.13
## 4.0.39
### Patch Changes
- 00114c5: feat: expose IDGenerator and createIdGenerator
- 00114c5: feat (ui): generate and forward message ids for response messages
- Updated dependencies [00114c5]
- Updated dependencies [00114c5]
- @ai-sdk/provider-utils@2.0.8
- @ai-sdk/ui-utils@1.0.11
- @ai-sdk/react@1.0.12
## 4.0.38
### Patch Changes
- 0118fa7: fix (ai/core): handle empty tool invocation array in convertToCoreMessages
## 4.0.37
### Patch Changes
- 8304ed8: feat (ai/core): Add option `throwErrorForEmptyVectors` to cosineSimilarity
- ed28182: feat (ai/ui): add appendResponseMessages helper
## 4.0.36
### Patch Changes
- Updated dependencies [37f4510]
- @ai-sdk/ui-utils@1.0.10
- @ai-sdk/react@1.0.11
## 4.0.35
### Patch Changes
- 3491f78: feat (ai/core): support multiple stream text transforms
## 4.0.34
### Patch Changes
- 2495973: feat (ai/core): use openai compatible mode for json schema conversion
- 2495973: fix (ai/core): duplicate instead of using reference in json schema
- Updated dependencies [2495973]
- Updated dependencies [2495973]
- @ai-sdk/ui-utils@1.0.9
- @ai-sdk/react@1.0.10
## 4.0.33
### Patch Changes
- 5510ee7: feat (ai/core): add stopStream option to streamText transforms
## 4.0.32
### Patch Changes
- de66619: feat (ai/core): add tool call id to ToolExecution error
## 4.0.31
### Patch Changes
- Updated dependencies [90fb95a]
- Updated dependencies [e6dfef4]
- Updated dependencies [6636db6]
- @ai-sdk/provider-utils@2.0.7
- @ai-sdk/react@1.0.9
- @ai-sdk/ui-utils@1.0.8
## 4.0.30
### Patch Changes
- e4ce80c: fix (ai/core): prevent onFinish from masking stream errors
## 4.0.29
### Patch Changes
- a92f5f6: feat (ai/core): generate many images with parallel model calls
## 4.0.28
### Patch Changes
- 19a2ce7: feat (ai/core): add aspectRatio and seed options to generateImage
- 6337688: feat: change image generation errors to warnings
- 8b422ea: feat (ai/core): add caching to generated images
- Updated dependencies [19a2ce7]
- Updated dependencies [19a2ce7]
- Updated dependencies [6337688]
- @ai-sdk/provider@1.0.4
- @ai-sdk/provider-utils@2.0.6
- @ai-sdk/ui-utils@1.0.7
- @ai-sdk/react@1.0.8
## 4.0.27
### Patch Changes
- a56734f: feat (ai/core): export simulateReadableStream in ai package
- 9589601: feat (ai/core): support null delay in smoothStream
- e3cc23a: feat (ai/core): support regexp chunking pattern in smoothStream
- e463e73: feat (ai/core): support skipping delays in simulateReadableStream
## 4.0.26
### Patch Changes
- a8f3242: feat (ai/core): add line chunking mode to smoothStream
## 4.0.25
### Patch Changes
- 0823899: fix (ai/core): throw error when accessing output when no output is defined in generateText (breaking/experimental)
## 4.0.24
### Patch Changes
- ae0485b: feat (ai/core): add experimental output setting to streamText
## 4.0.23
### Patch Changes
- bc4cd19: feat (ai/core): consolidate whitespace in smooth stream
## 4.0.22
### Patch Changes
- Updated dependencies [5ed5e45]
- @ai-sdk/provider-utils@2.0.5
- @ai-sdk/provider@1.0.3
- @ai-sdk/react@1.0.7
- @ai-sdk/ui-utils@1.0.6
## 4.0.21
### Patch Changes
- a8669a2: fix (ai/core): prefer auto-detected image mimetype
- 6fb3e91: fix (ai/core): include type in generateText toolResults result property.
## 4.0.20
### Patch Changes
- da9d240: fix (ai/core): suppress errors caused by writing to closed stream
- 6f1bfde: fix (ai/core): invoke streamText tool call repair when tool cannot be found
## 4.0.19
### Patch Changes
- c3a6065: fix (ai/core): apply transform before callbacks and resolvables
## 4.0.18
### Patch Changes
- 304e6d3: feat (ai/core): standardize generateObject, streamObject, and output errors to NoObjectGeneratedError
- 304e6d3: feat (ai/core): add additional information to NoObjectGeneratedError
## 4.0.17
### Patch Changes
- 54bbf21: fix (ai/core): change streamText.experimental_transform signature to support tool type inference
## 4.0.16
### Patch Changes
- e3fac3f: fix (ai/core): change smoothStream default delay to 10ms
## 4.0.15
### Patch Changes
- cc16a83: feat (ai/core): add smoothStream helper
- cc16a83: feat (ai/core): add experimental transform option to streamText
## 4.0.14
### Patch Changes
- 09a9cab: feat (ai/core): add experimental generateImage function
- Updated dependencies [09a9cab]
- @ai-sdk/provider@1.0.2
- @ai-sdk/provider-utils@2.0.4
- @ai-sdk/ui-utils@1.0.5
- @ai-sdk/react@1.0.6
## 4.0.13
### Patch Changes
- 9f32213: feat (ai/core): add experimental tool call repair
## 4.0.12
### Patch Changes
- 5167bec: fix (ai/core): forward streamText errors as error parts
- 0984f0b: feat (ai/core): add ToolExecutionError type
- Updated dependencies [0984f0b]
- @ai-sdk/provider-utils@2.0.3
- @ai-sdk/react@1.0.5
- @ai-sdk/ui-utils@1.0.4
## 4.0.11
### Patch Changes
- Updated dependencies [953469c]
- Updated dependencies [a3dd2ed]
- @ai-sdk/ui-utils@1.0.3
- @ai-sdk/react@1.0.4
## 4.0.10
### Patch Changes
- 913872d: fix (ai/core): track promise from async createDataStream.execute
## 4.0.9
### Patch Changes
- fda9695: feat (ai/core): reworked data stream management
## 4.0.8
### Patch Changes
- a803d76: feat (ai/core): pass toolCallId option into tool execute function
## 4.0.7
### Patch Changes
- 5b4f07b: fix (ai/core): change default error message for data streams to "An error occurred."
## 4.0.6
### Patch Changes
- fc18132: feat (ai/core): experimental output for generateText
- 2779f6d: fix (ai/core): do not send maxRetries into providers
## 4.0.5
### Patch Changes
- Updated dependencies [630ac31]
- @ai-sdk/react@1.0.3
## 4.0.4
### Patch Changes
- 6ff6689: fix (ai): trigger onFinal when stream adapter finishes
- 6ff6689: chore (ai): deprecate onCompletion (stream callbacks)
## 4.0.3
### Patch Changes
- Updated dependencies [88b364b]
- Updated dependencies [b446ae5]
- @ai-sdk/ui-utils@1.0.2
- @ai-sdk/provider@1.0.1
- @ai-sdk/react@1.0.2
- @ai-sdk/provider-utils@2.0.2
## 4.0.2
### Patch Changes
- Updated dependencies [c3ab5de]
- @ai-sdk/provider-utils@2.0.1
- @ai-sdk/react@1.0.1
- @ai-sdk/ui-utils@1.0.1
## 4.0.1
### Patch Changes
- b117255: feat (ai/core): add messages to tool call options
## 4.0.0
### Major Changes
- 4e38b38: chore (ai): remove LanguageModelResponseMetadataWithHeaders type
- 8bf5756: chore: remove legacy function/tool calling
- f0cb69d: chore (ai/core): remove experimental function exports
- da8c609: chore (ai): remove Tokens RSC helper
- cbab571: chore (ai): remove ExperimentalXXXMessage types
- b469a7e: chore: remove isXXXError methods
- 54cb888: chore (ai): remove experimental_StreamData export
- 4d61295: chore (ai): remove streamToResponse and streamingTextResponse
- 9a3d741: chore (ai): remove ExperimentalTool export
- 064257d: chore (ai/core): rename simulateReadableStream values parameter to chunks
- 60e69ed: chore (ai/core): remove ai-stream related methods from streamText
- a4f8ce9: chore (ai): AssistantResponse cleanups
- d3ae4f6: chore (ui/react): remove useObject setInput helper
- 7264b0a: chore (ai): remove responseMessages property from streamText/generateText result
- b801982: chore (ai/core): remove init option from streamText result methods
- f68d7b1: chore (ai/core): streamObject returns result immediately (no Promise)
- 6090cea: chore (ai): remove rawResponse from generate/stream result objects
- 073f282: chore (ai): remove AIStream and related exports
- 1c58337: chore (ai): remove 2.x prompt helpers
- a40a93d: chore (ai/ui): remove vue, svelte, solid re-export and dependency
- a7ad35a: chore: remove legacy providers & rsc render
- c0ddc24: chore (ai): remove toJSON method from AI SDK errors
- 007cb81: chore (ai): change `streamText` warnings result to Promise
- effbce3: chore (ai): remove responseMessage from streamText onFinish callback
- 545d133: chore (ai): remove deprecated roundtrip settings from streamText / generateText
- 7e89ccb: chore: remove nanoid export
- f967199: chore (ai/core): streamText returns result immediately (no Promise)
- 62d08fd: chore (ai): remove TokenUsage, CompletionTokenUsage, and EmbeddingTokenUsage types
- e5d2ce8: chore (ai): remove deprecated provider registry exports
- 70ce742: chore (ai): remove experimental_continuationSteps option
- 2f09717: chore (ai): remove deprecated telemetry data
- 0827bf9: chore (ai): remove LangChain adapter `toAIStream` method
### Patch Changes
- dce4158: chore (dependencies): update eventsource-parser to 3.0.0
- f0ec721: chore (ai): remove openai peer dependency
- f9bb30c: chore (ai): remove unnecessary dev dependencies
- b053413: chore (ui): refactorings & README update
- Updated dependencies [e117b54]
- Updated dependencies [8bf5756]
- Updated dependencies [b469a7e]
- Updated dependencies [79c6dd9]
- Updated dependencies [9f81e66]
- Updated dependencies [70f28f6]
- Updated dependencies [dce4158]
- Updated dependencies [d3ae4f6]
- Updated dependencies [68d30e9]
- Updated dependencies [7814c4b]
- Updated dependencies [ca3e586]
- Updated dependencies [c0ddc24]
- Updated dependencies [fe4f109]
- Updated dependencies [84edae5]
- Updated dependencies [b1da952]
- Updated dependencies [04d3747]
- Updated dependencies [dce4158]
- Updated dependencies [7e89ccb]
- Updated dependencies [8426f55]
- Updated dependencies [db46ce5]
- Updated dependencies [b053413]
- @ai-sdk/react@1.0.0
- @ai-sdk/ui-utils@1.0.0
- @ai-sdk/provider-utils@2.0.0
- @ai-sdk/provider@1.0.0
## 4.0.0-canary.13
### Major Changes
- 064257d: chore (ai/core): rename simulateReadableStream values parameter to chunks
### Patch Changes
- Updated dependencies [79c6dd9]
- Updated dependencies [04d3747]
- @ai-sdk/react@1.0.0-canary.9
- @ai-sdk/ui-utils@1.0.0-canary.9
## 4.0.0-canary.12
### Patch Changes
- b053413: chore (ui): refactorings & README update
- Updated dependencies [b053413]
- @ai-sdk/ui-utils@1.0.0-canary.8
- @ai-sdk/react@1.0.0-canary.8
## 4.0.0-canary.11
### Major Changes
- f68d7b1: chore (ai/core): streamObject returns result immediately (no Promise)
- f967199: chore (ai/core): streamText returns result immediately (no Promise)
## 4.0.0-canary.10
### Major Changes
- effbce3: chore (ai): remove responseMessage from streamText onFinish callback
### Patch Changes
- Updated dependencies [fe4f109]
- @ai-sdk/ui-utils@1.0.0-canary.7
- @ai-sdk/react@1.0.0-canary.7
## 4.0.0-canary.9
### Patch Changes
- f0ec721: chore (ai): remove openai peer dependency
## 4.0.0-canary.8
### Major Changes
- 007cb81: chore (ai): change `streamText` warnings result to Promise
### Patch Changes
- Updated dependencies [70f28f6]
- @ai-sdk/ui-utils@1.0.0-canary.6
- @ai-sdk/react@1.0.0-canary.6
## 4.0.0-canary.7
### Major Changes
- 4e38b38: chore (ai): remove LanguageModelResponseMetadataWithHeaders type
- 54cb888: chore (ai): remove experimental_StreamData export
- 9a3d741: chore (ai): remove ExperimentalTool export
- a4f8ce9: chore (ai): AssistantResponse cleanups
- 7264b0a: chore (ai): remove responseMessages property from streamText/generateText result
- 62d08fd: chore (ai): remove TokenUsage, CompletionTokenUsage, and EmbeddingTokenUsage types
- e5d2ce8: chore (ai): remove deprecated provider registry exports
- 70ce742: chore (ai): remove experimental_continuationSteps option
- 0827bf9: chore (ai): remove LangChain adapter `toAIStream` method
## 4.0.0-canary.6
### Major Changes
- b801982: chore (ai/core): remove init option from streamText result methods
### Patch Changes
- f9bb30c: chore (ai): remove unnecessary dev dependencies
## 4.0.0-canary.5
### Major Changes
- 4d61295: chore (ai): remove streamToResponse and streamingTextResponse
- d3ae4f6: chore (ui/react): remove useObject setInput helper
- 6090cea: chore (ai): remove rawResponse from generate/stream result objects
- 2f09717: chore (ai): remove deprecated telemetry data
### Patch Changes
- Updated dependencies [9f81e66]
- Updated dependencies [d3ae4f6]
- Updated dependencies [8426f55]
- @ai-sdk/ui-utils@1.0.0-canary.5
- @ai-sdk/react@1.0.0-canary.5
- @ai-sdk/provider-utils@2.0.0-canary.3
## 4.0.0-canary.4
### Major Changes
- f0cb69d: chore (ai/core): remove experimental function exports
- da8c609: chore (ai): remove Tokens RSC helper
- cbab571: chore (ai): remove ExperimentalXXXMessage types
- 60e69ed: chore (ai/core): remove ai-stream related methods from streamText
- 073f282: chore (ai): remove AIStream and related exports
- 545d133: chore (ai): remove deprecated roundtrip settings from streamText / generateText
### Patch Changes
- dce4158: chore (dependencies): update eventsource-parser to 3.0.0
- Updated dependencies [dce4158]
- Updated dependencies [ca3e586]
- Updated dependencies [dce4158]
- @ai-sdk/provider-utils@2.0.0-canary.2
- @ai-sdk/react@1.0.0-canary.4
- @ai-sdk/ui-utils@1.0.0-canary.4
## 4.0.0-canary.3
### Patch Changes
- Updated dependencies [68d30e9]
- Updated dependencies [b1da952]
- @ai-sdk/react@1.0.0-canary.3
- @ai-sdk/provider-utils@2.0.0-canary.1
- @ai-sdk/ui-utils@1.0.0-canary.3
## 4.0.0-canary.2
### Major Changes
- b469a7e: chore: remove isXXXError methods
- c0ddc24: chore (ai): remove toJSON method from AI SDK errors
### Patch Changes
- Updated dependencies [e117b54]
- Updated dependencies [b469a7e]
- Updated dependencies [7814c4b]
- Updated dependencies [c0ddc24]
- Updated dependencies [db46ce5]
- @ai-sdk/react@1.0.0-canary.2
- @ai-sdk/provider-utils@2.0.0-canary.0
- @ai-sdk/provider@1.0.0-canary.0
- @ai-sdk/ui-utils@1.0.0-canary.2
## 4.0.0-canary.1
### Major Changes
- 8bf5756: chore: remove legacy function/tool calling
### Patch Changes
- 1c58337: chore (ai): remove 2.x prompt helpers
- Updated dependencies [8bf5756]
- @ai-sdk/ui-utils@1.0.0-canary.1
- @ai-sdk/react@1.0.0-canary.1
## 4.0.0-canary.0
### Major Changes
- a40a93d: chore (ai/ui): remove vue, svelte, solid re-export and dependency
### Patch Changes
- a7ad35a: chore: remove legacy providers & rsc render
- 7e89ccb: chore: remove nanoid export
- Updated dependencies [84edae5]
- Updated dependencies [7e89ccb]
- @ai-sdk/react@1.0.0-canary.0
- @ai-sdk/ui-utils@1.0.0-canary.0
## 3.4.33
### Patch Changes
- ac380e3: fix (provider/anthropic): continuation mode with 3+ steps
## 3.4.32
### Patch Changes
- 6bb9e51: fix (ai/core): expose response.messages in streamText
## 3.4.31
### Patch Changes
- Updated dependencies [2dfb93e]
- @ai-sdk/react@0.0.70
## 3.4.30
### Patch Changes
- Updated dependencies [a85c965]
- @ai-sdk/ui-utils@0.0.50
- @ai-sdk/react@0.0.69
- @ai-sdk/solid@0.0.54
- @ai-sdk/svelte@0.0.57
- @ai-sdk/vue@0.0.59
## 3.4.29
### Patch Changes
- 54b56f7: feat (ai/core): send tool and tool choice telemetry data
## 3.4.28
### Patch Changes
- 29f1390: feat (ai/test): add simulateReadableStream helper
## 3.4.27
### Patch Changes
- fa772ae: feat (ai/core): automatically convert ui messages to core messages
## 3.4.26
### Patch Changes
- 57f39ea: feat (ai/core): support multi-modal tool results in convertToCoreMessages
## 3.4.25
### Patch Changes
- 6e0fa1c: fix (ai/core): wait for tool results to arrive before sending finish event
## 3.4.24
### Patch Changes
- d92fd9f: feat (ui/svelte): support Svelte 5 peer dependency
- Updated dependencies [d92fd9f]
- @ai-sdk/svelte@0.0.56
## 3.4.23
### Patch Changes
- 8301e41: fix (ai/react): update React peer dependency version to allow rc releases.
- Updated dependencies [8301e41]
- @ai-sdk/react@0.0.68
## 3.4.22
### Patch Changes
- Updated dependencies [3bf8da0]
- @ai-sdk/ui-utils@0.0.49
- @ai-sdk/react@0.0.67
- @ai-sdk/solid@0.0.53
- @ai-sdk/svelte@0.0.55
- @ai-sdk/vue@0.0.58
## 3.4.21
### Patch Changes
- 3954471: (experimental) fix passing "experimental_toToolResultContent" into PoolResultPart
## 3.4.20
### Patch Changes
- aa98cdb: chore: more flexible dependency versioning
- 1486128: feat: add supportsUrl to language model specification
- 3b1b69a: feat: provider-defined tools
- 85b98da: revert fix (ai/core): handle tool calls without results in message conversion
- 7ceed77: feat (ai/core): expose response message for each step
- 811a317: feat (ai/core): multi-part tool results (incl. images)
- Updated dependencies [aa98cdb]
- Updated dependencies [1486128]
- Updated dependencies [7b937c5]
- Updated dependencies [3b1b69a]
- Updated dependencies [811a317]
- @ai-sdk/provider-utils@1.0.22
- @ai-sdk/provider@0.0.26
- @ai-sdk/ui-utils@0.0.48
- @ai-sdk/svelte@0.0.54
- @ai-sdk/react@0.0.66
- @ai-sdk/vue@0.0.57
- @ai-sdk/solid@0.0.52
## 3.4.19
### Patch Changes
- b9b0d7b: feat (ai): access raw request body
- Updated dependencies [b9b0d7b]
- @ai-sdk/provider@0.0.25
- @ai-sdk/provider-utils@1.0.21
- @ai-sdk/ui-utils@0.0.47
- @ai-sdk/react@0.0.65
- @ai-sdk/solid@0.0.51
- @ai-sdk/svelte@0.0.53
- @ai-sdk/vue@0.0.56
## 3.4.18
### Patch Changes
- 95c67b4: fix (ai/core): handle tool calls without results in message conversion
## 3.4.17
### Patch Changes
- e4ff512: fix (core): prevent unnecessary input/output serialization when telemetry is not enabled
## 3.4.16
### Patch Changes
- 01dcc44: feat (ai/core): add experimental activeTools option to generateText and streamText
## 3.4.15
### Patch Changes
- Updated dependencies [98a3b08]
- @ai-sdk/react@0.0.64
## 3.4.14
### Patch Changes
- e930f40: feat (ai/core): expose core tool result and tool call types
## 3.4.13
### Patch Changes
- fc39158: fix (ai/core): add abortSignal to tool helper function
## 3.4.12
### Patch Changes
- a23da5b: feat (ai/core): forward abort signal to tools
## 3.4.11
### Patch Changes
- caedcda: feat (ai/ui): add setData helper to useChat
- Updated dependencies [caedcda]
- @ai-sdk/svelte@0.0.52
- @ai-sdk/react@0.0.63
- @ai-sdk/solid@0.0.50
- @ai-sdk/vue@0.0.55
## 3.4.10
### Patch Changes
- 0b557d7: feat (ai/core): add tracer option to telemetry settings
- 44f6bc5: feat (ai/core): expose StepResult type
## 3.4.9
### Patch Changes
- d347538: fix (ai/core): export FilePart interface
## 3.4.8
### Patch Changes
- Updated dependencies [b5f577e]
- @ai-sdk/vue@0.0.54
## 3.4.7
### Patch Changes
- db04700: feat (core): support converting attachments to file parts
- 988707c: feat (ai/core): automatically download files from urls
## 3.4.6
### Patch Changes
- d595d0d: feat (ai/core): file content parts
- Updated dependencies [d595d0d]
- @ai-sdk/provider@0.0.24
- @ai-sdk/provider-utils@1.0.20
- @ai-sdk/ui-utils@0.0.46
- @ai-sdk/react@0.0.62
- @ai-sdk/solid@0.0.49
- @ai-sdk/svelte@0.0.51
- @ai-sdk/vue@0.0.53
## 3.4.5
### Patch Changes
- cd77c5d: feat (ai/core): add isContinued to steps
- Updated dependencies [cd77c5d]
- @ai-sdk/ui-utils@0.0.45
- @ai-sdk/react@0.0.61
- @ai-sdk/solid@0.0.48
- @ai-sdk/svelte@0.0.50
- @ai-sdk/vue@0.0.52
## 3.4.4
### Patch Changes
- 4db074b: fix (ai/core): correct whitespace in generateText continueSteps
- 1297e1b: fix (ai/core): correct whitespace in streamText continueSteps
## 3.4.3
### Patch Changes
- b270ae3: feat (ai/core): streamText continueSteps (experimental)
- b270ae3: chore (ai/core): rename generateText continuationSteps to continueSteps
## 3.4.2
### Patch Changes
- e6c7e98: feat (ai/core): add continuationSteps to generateText
## 3.4.1
### Patch Changes
- Updated dependencies [7e7104f]
- @ai-sdk/react@0.0.60
## 3.4.0
### Minor Changes
- c0cea03: release (ai): 3.4
## 3.3.44
### Patch Changes
- Updated dependencies [d3933e0]
- @ai-sdk/vue@0.0.51
## 3.3.43
### Patch Changes
- fea6bec: fix (ai/core): support tool calls without arguments
## 3.3.42
### Patch Changes
- de37aee: feat (ai): Add support for LlamaIndex
## 3.3.41
### Patch Changes
- Updated dependencies [692e265]
- @ai-sdk/vue@0.0.50
## 3.3.40
### Patch Changes
- a91c308: feat (ai/core): add responseMessages to streamText
## 3.3.39
### Patch Changes
- 33cf3e1: feat (ai/core): add providerMetadata to StepResult
- 17ee757: feat (ai/core): add onStepFinish callback to generateText
## 3.3.38
### Patch Changes
- 83da52c: feat (ai/core): add onStepFinish callback to streamText
## 3.3.37
### Patch Changes
- Updated dependencies [273f696]
- @ai-sdk/provider-utils@1.0.19
- @ai-sdk/react@0.0.59
- @ai-sdk/solid@0.0.47
- @ai-sdk/svelte@0.0.49
- @ai-sdk/ui-utils@0.0.44
- @ai-sdk/vue@0.0.49
## 3.3.36
### Patch Changes
- a3882f5: feat (ai/core): add steps property to streamText result and onFinish callback
- 1f590ef: chore (ai): rename roundtrips to steps
- 7e82d36: fix (ai/core): pass topK to providers
- Updated dependencies [54862e4]
- Updated dependencies [1f590ef]
- @ai-sdk/react@0.0.58
- @ai-sdk/ui-utils@0.0.43
- @ai-sdk/solid@0.0.46
- @ai-sdk/svelte@0.0.48
- @ai-sdk/vue@0.0.48
## 3.3.35
### Patch Changes
- 14210d5: feat (ai/core): add sendUsage information to streamText data stream methods
- Updated dependencies [14210d5]
- @ai-sdk/ui-utils@0.0.42
- @ai-sdk/react@0.0.57
- @ai-sdk/solid@0.0.45
- @ai-sdk/svelte@0.0.47
- @ai-sdk/vue@0.0.47
## 3.3.34
### Patch Changes
- a0403d6: feat (react): support sending attachments using append
- 678449a: feat (ai/core): export test helpers
- ff22fac: fix (ai/rsc): streamUI onFinish is called when tool calls have finished
- Updated dependencies [a0403d6]
- @ai-sdk/react@0.0.56
## 3.3.33
### Patch Changes
- cbddc83: fix (ai/core): filter out empty text parts
## 3.3.32
### Patch Changes
- ce7a4af: feat (ai/core): support providerMetadata in functions
## 3.3.31
### Patch Changes
- 561fd7e: feat (ai/core): add output: enum to generateObject
## 3.3.30
### Patch Changes
- 6ee1f8e: feat (ai/core): add toDataStream to streamText result
## 3.3.29
### Patch Changes
- 1e3dfd2: feat (ai/core): enhance pipeToData/TextStreamResponse methods
## 3.3.28
### Patch Changes
- db61c53: feat (ai/core): middleware support
## 3.3.27
### Patch Changes
- 03313cd: feat (ai): expose response id, response model, response timestamp in telemetry and api
- 3be7c1c: fix (provider/anthropic): support prompt caching on assistant messages
- Updated dependencies [03313cd]
- Updated dependencies [3be7c1c]
- @ai-sdk/provider-utils@1.0.18
- @ai-sdk/provider@0.0.23
- @ai-sdk/react@0.0.55
- @ai-sdk/solid@0.0.44
- @ai-sdk/svelte@0.0.46
- @ai-sdk/ui-utils@0.0.41
- @ai-sdk/vue@0.0.46
## 3.3.26
### Patch Changes
- Updated dependencies [4ab883f]
- @ai-sdk/react@0.0.54
## 3.3.25
### Patch Changes
- 4f1530f: feat (ai/core): add OpenTelemetry Semantic Conventions for GenAI operations to v1.27.0 of standard
- dad775f: feat (ai/core): add finish event and avg output tokens per second (telemetry)
## 3.3.24
### Patch Changes
- d87a655: fix (ai/core): provide fallback when globalThis.performance is not available
## 3.3.23
### Patch Changes
- b55e6f7: fix (ai/core): streamObject text stream in array mode must not include elements: prefix.
## 3.3.22
### Patch Changes
- a5a56fd: fix (ai/core): only send roundtrip-finish event after async tool calls are done
## 3.3.21
### Patch Changes
- aa2dc58: feat (ai/core): add maxToolRoundtrips to streamText
- Updated dependencies [aa2dc58]
- @ai-sdk/ui-utils@0.0.40
- @ai-sdk/react@0.0.53
- @ai-sdk/solid@0.0.43
- @ai-sdk/svelte@0.0.45
- @ai-sdk/vue@0.0.45
## 3.3.20
### Patch Changes
- 7807677: fix (rsc): Deep clone currentState in getMutableState()
## 3.3.19
### Patch Changes
- 7235de0: fix (ai/core): convertToCoreMessages accepts Message[]
## 3.3.18
### Patch Changes
- 9e3b5a5: feat (ai/core): add experimental_customProvider
- 26515cb: feat (ai/provider): introduce ProviderV1 specification
- Updated dependencies [26515cb]
- @ai-sdk/provider@0.0.22
- @ai-sdk/provider-utils@1.0.17
- @ai-sdk/ui-utils@0.0.39
- @ai-sdk/react@0.0.52
- @ai-sdk/solid@0.0.42
- @ai-sdk/svelte@0.0.44
- @ai-sdk/vue@0.0.44
## 3.3.17
### Patch Changes
- d151349: feat (ai/core): array output for generateObject / streamObject
- Updated dependencies [d151349]
- @ai-sdk/ui-utils@0.0.38
- @ai-sdk/react@0.0.51
- @ai-sdk/solid@0.0.41
- @ai-sdk/svelte@0.0.43
- @ai-sdk/vue@0.0.43
## 3.3.16
### Patch Changes
- 09f895f: feat (ai/core): no-schema output for generateObject / streamObject
- Updated dependencies [09f895f]
- @ai-sdk/provider-utils@1.0.16
- @ai-sdk/react@0.0.50
- @ai-sdk/solid@0.0.40
- @ai-sdk/svelte@0.0.42
- @ai-sdk/ui-utils@0.0.37
- @ai-sdk/vue@0.0.42
## 3.3.15
### Patch Changes
- b5a82b7: chore (ai): update zod-to-json-schema to 3.23.2
- Updated dependencies [b5a82b7]
- @ai-sdk/ui-utils@0.0.36
- @ai-sdk/react@0.0.49
- @ai-sdk/solid@0.0.39
- @ai-sdk/svelte@0.0.41
- @ai-sdk/vue@0.0.41
## 3.3.14
### Patch Changes
- Updated dependencies [d67fa9c]
- @ai-sdk/provider-utils@1.0.15
- @ai-sdk/react@0.0.48
- @ai-sdk/solid@0.0.38
- @ai-sdk/svelte@0.0.40
- @ai-sdk/ui-utils@0.0.35
- @ai-sdk/vue@0.0.40
## 3.3.13
### Patch Changes
- 412f943: fix (ai/core): make Buffer validation optional for environments without buffer
## 3.3.12
### Patch Changes
- f2c025e: feat (ai/core): prompt validation
- Updated dependencies [f2c025e]
- @ai-sdk/provider@0.0.21
- @ai-sdk/provider-utils@1.0.14
- @ai-sdk/ui-utils@0.0.34
- @ai-sdk/react@0.0.47
- @ai-sdk/solid@0.0.37
- @ai-sdk/svelte@0.0.39
- @ai-sdk/vue@0.0.39
## 3.3.11
### Patch Changes
- 03eb0f4: feat (ai/core): add "ai.operationId" telemetry attribute
- 099db96: feat (ai/core): add msToFirstChunk telemetry data
- Updated dependencies [b6c1dee]
- @ai-sdk/react@0.0.46
## 3.3.10
### Patch Changes
- Updated dependencies [04084a3]
- @ai-sdk/vue@0.0.38
## 3.3.9
### Patch Changes
- 6ac355e: feat (provider/anthropic): add cache control support
- b56dee1: chore (ai): deprecate prompt helpers
- Updated dependencies [6ac355e]
- @ai-sdk/provider@0.0.20
- @ai-sdk/provider-utils@1.0.13
- @ai-sdk/ui-utils@0.0.33
- @ai-sdk/react@0.0.45
- @ai-sdk/solid@0.0.36
- @ai-sdk/svelte@0.0.38
- @ai-sdk/vue@0.0.37
## 3.3.8
### Patch Changes
- Updated dependencies [dd712ac]
- @ai-sdk/provider-utils@1.0.12
- @ai-sdk/ui-utils@0.0.32
- @ai-sdk/react@0.0.44
- @ai-sdk/solid@0.0.35
- @ai-sdk/svelte@0.0.37
- @ai-sdk/vue@0.0.36
## 3.3.7
### Patch Changes
- eccbd8e: feat (ai/core): add onChunk callback to streamText
- Updated dependencies [dd4a0f5]
- @ai-sdk/provider@0.0.19
- @ai-sdk/provider-utils@1.0.11
- @ai-sdk/ui-utils@0.0.31
- @ai-sdk/react@0.0.43
- @ai-sdk/solid@0.0.34
- @ai-sdk/svelte@0.0.36
- @ai-sdk/vue@0.0.35
## 3.3.6
### Patch Changes
- e9c891d: feat (ai/react): useObject supports non-Zod schemas
- 3719e8a: chore (ai/core): provider registry code improvements
- Updated dependencies [e9c891d]
- Updated dependencies [4bd27a9]
- Updated dependencies [845754b]
- @ai-sdk/ui-utils@0.0.30
- @ai-sdk/react@0.0.42
- @ai-sdk/provider-utils@1.0.10
- @ai-sdk/provider@0.0.18
- @ai-sdk/solid@0.0.33
- @ai-sdk/svelte@0.0.35
- @ai-sdk/vue@0.0.34
## 3.3.5
### Patch Changes
- 9ada023: feat (ai/core): mask data stream error messages with streamText
- Updated dependencies [e5b58f3]
- @ai-sdk/ui-utils@0.0.29
- @ai-sdk/react@0.0.41
- @ai-sdk/solid@0.0.32
- @ai-sdk/svelte@0.0.34
- @ai-sdk/vue@0.0.33
## 3.3.4
### Patch Changes
- 029af4c: feat (ai/core): support schema name & description in generateObject & streamObject
- 3806c0c: chore (ai/ui): increase stream data warning timeout to 15 seconds
- db0118a: feat (ai/core): export Schema type
- Updated dependencies [029af4c]
- @ai-sdk/provider@0.0.17
- @ai-sdk/provider-utils@1.0.9
- @ai-sdk/ui-utils@0.0.28
- @ai-sdk/react@0.0.40
- @ai-sdk/solid@0.0.31
- @ai-sdk/svelte@0.0.33
- @ai-sdk/vue@0.0.32
## 3.3.3
### Patch Changes
- d58517b: feat (ai/openai): structured outputs
- Updated dependencies [d58517b]
- @ai-sdk/provider@0.0.16
- @ai-sdk/provider-utils@1.0.8
- @ai-sdk/ui-utils@0.0.27
- @ai-sdk/react@0.0.39
- @ai-sdk/solid@0.0.30
- @ai-sdk/svelte@0.0.32
- @ai-sdk/vue@0.0.31
## 3.3.2
### Patch Changes
- Updated dependencies [96aed25]
- @ai-sdk/provider@0.0.15
- @ai-sdk/provider-utils@1.0.7
- @ai-sdk/ui-utils@0.0.26
- @ai-sdk/react@0.0.38
- @ai-sdk/solid@0.0.29
- @ai-sdk/svelte@0.0.31
- @ai-sdk/vue@0.0.30
## 3.3.1
### Patch Changes
- 9614584: fix (ai/core): use Symbol.for
- 0762a22: feat (ai/core): support zod transformers in generateObject & streamObject
- Updated dependencies [9614584]
- Updated dependencies [0762a22]
- @ai-sdk/provider-utils@1.0.6
- @ai-sdk/react@0.0.37
- @ai-sdk/solid@0.0.28
- @ai-sdk/svelte@0.0.30
- @ai-sdk/ui-utils@0.0.25
- @ai-sdk/vue@0.0.29
## 3.3.0
### Minor Changes
- dbc3afb7: chore (ai): release AI SDK 3.3
### Patch Changes
- b9827186: feat (ai/core): update operation.name telemetry attribute to include function id and detailed name
## 3.2.45
### Patch Changes
- Updated dependencies [5be25124]
- @ai-sdk/ui-utils@0.0.24
- @ai-sdk/react@0.0.36
- @ai-sdk/solid@0.0.27
- @ai-sdk/svelte@0.0.29
- @ai-sdk/vue@0.0.28
## 3.2.44
### Patch Changes
- Updated dependencies [a147d040]
- @ai-sdk/react@0.0.35
## 3.2.43
### Patch Changes
- Updated dependencies [b68fae4f]
- @ai-sdk/react@0.0.34
## 3.2.42
### Patch Changes
- f63c99e7: feat (ai/core): record OpenTelemetry gen_ai attributes
- Updated dependencies [fea7b604]
- @ai-sdk/ui-utils@0.0.23
- @ai-sdk/react@0.0.33
- @ai-sdk/solid@0.0.26
- @ai-sdk/svelte@0.0.28
- @ai-sdk/vue@0.0.27
## 3.2.41
### Patch Changes
- a12044c7: feat (ai/core): add recordInputs / recordOutputs setting to telemetry options
- Updated dependencies [1d93d716]
- @ai-sdk/ui-utils@0.0.22
- @ai-sdk/react@0.0.32
- @ai-sdk/solid@0.0.25
- @ai-sdk/svelte@0.0.27
- @ai-sdk/vue@0.0.26
## 3.2.40
### Patch Changes
- f56b7e66: feat (ai/ui): add toDataStreamResponse to LangchainAdapter.
## 3.2.39
### Patch Changes
- b694f2f9: feat (ai/svelte): add tool calling support to useChat
- Updated dependencies [b694f2f9]
- @ai-sdk/svelte@0.0.26
## 3.2.38
### Patch Changes
- 5c4b8cfc: chore (ai/core): rename ai stream methods to data stream (in streamText, LangChainAdapter).
- c450fcf7: feat (ui): invoke useChat onFinish with finishReason and tokens
- e4a1719f: chore (ai/ui): rename streamMode to streamProtocol
- 10158bf2: fix (ai/core): generateObject.doGenerate sets object telemetry attribute
- Updated dependencies [c450fcf7]
- Updated dependencies [e4a1719f]
- @ai-sdk/ui-utils@0.0.21
- @ai-sdk/svelte@0.0.25
- @ai-sdk/react@0.0.31
- @ai-sdk/solid@0.0.24
- @ai-sdk/vue@0.0.25
## 3.2.37
### Patch Changes
- b2bee4c5: fix (ai/ui): send data, body, headers in useChat().reload
- Updated dependencies [b2bee4c5]
- @ai-sdk/svelte@0.0.24
- @ai-sdk/react@0.0.30
- @ai-sdk/solid@0.0.23
## 3.2.36
### Patch Changes
- a8d1c9e9: feat (ai/core): parallel image download
- cfa360a8: feat (ai/core): add telemetry support to embedMany function.
- 49808ca5: feat (ai/core): add telemetry to streamObject
- Updated dependencies [a8d1c9e9]
- @ai-sdk/provider-utils@1.0.5
- @ai-sdk/provider@0.0.14
- @ai-sdk/react@0.0.29
- @ai-sdk/svelte@0.0.23
- @ai-sdk/ui-utils@0.0.20
- @ai-sdk/vue@0.0.24
- @ai-sdk/solid@0.0.22
## 3.2.35
### Patch Changes
- 1be014b7: feat (ai/core): add telemetry support for embed function.
- 4f88248f: feat (core): support json schema
- 0d545231: chore (ai/svelte): change sswr into optional peer dependency
- Updated dependencies [4f88248f]
- @ai-sdk/provider-utils@1.0.4
- @ai-sdk/react@0.0.28
- @ai-sdk/svelte@0.0.22
- @ai-sdk/ui-utils@0.0.19
- @ai-sdk/vue@0.0.23
- @ai-sdk/solid@0.0.21
## 3.2.34
### Patch Changes
- 2b9da0f0: feat (core): support stopSequences setting.
- a5b58845: feat (core): support topK setting
- 420f170f: chore (ai/core): use interfaces for core function results
- 13b27ec6: chore (ai/core): remove grammar mode
- 644f6582: feat (ai/core): add telemetry to generateObject
- Updated dependencies [2b9da0f0]
- Updated dependencies [a5b58845]
- Updated dependencies [4aa8deb3]
- Updated dependencies [13b27ec6]
- @ai-sdk/provider@0.0.13
- @ai-sdk/provider-utils@1.0.3
- @ai-sdk/react@0.0.27
- @ai-sdk/svelte@0.0.21
- @ai-sdk/ui-utils@0.0.18
- @ai-sdk/solid@0.0.20
- @ai-sdk/vue@0.0.22
## 3.2.33
### Patch Changes
- 4b2c09d9: feat (ai/ui): add mutator function support to useChat / setMessages
- 281e7662: chore: add description to ai package
- Updated dependencies [f63829fe]
- Updated dependencies [4b2c09d9]
- @ai-sdk/ui-utils@0.0.17
- @ai-sdk/svelte@0.0.20
- @ai-sdk/react@0.0.26
- @ai-sdk/solid@0.0.19
- @ai-sdk/vue@0.0.21
## 3.2.32
### Patch Changes
- Updated dependencies [5b7b3bbe]
- @ai-sdk/ui-utils@0.0.16
- @ai-sdk/react@0.0.25
- @ai-sdk/solid@0.0.18
- @ai-sdk/svelte@0.0.19
- @ai-sdk/vue@0.0.20
## 3.2.31
### Patch Changes
- b86af092: feat (ai/core): add langchain stream event v2 support to LangChainAdapter
## 3.2.30
### Patch Changes
- Updated dependencies [19c3d50f]
- @ai-sdk/react@0.0.24
- @ai-sdk/vue@0.0.19
## 3.2.29
### Patch Changes
- e710b388: fix (ai/core): race condition in mergeStreams
- 6078a690: feat (ai/core): introduce stream data support in toAIStreamResponse
## 3.2.28
### Patch Changes
- 68d1f78c: fix (ai/core): do not construct object promise in streamObject result until requested
- f0bc1e79: feat (ai/ui): add system message support to convertToCoreMessages
- 1f67fe49: feat (ai/ui): stream tool calls with streamText and useChat
- Updated dependencies [1f67fe49]
- @ai-sdk/ui-utils@0.0.15
- @ai-sdk/react@0.0.23
- @ai-sdk/solid@0.0.17
- @ai-sdk/svelte@0.0.18
- @ai-sdk/vue@0.0.18
## 3.2.27
### Patch Changes
- 811f4493: fix (ai/core): generateText token usage is sum over all roundtrips
## 3.2.26
### Patch Changes
- 8f545ce9: fix (ai/core): forward request headers in generateObject and streamObject
## 3.2.25
### Patch Changes
- 99ddbb74: feat (ai/react): add experimental support for managing attachments to useChat
- Updated dependencies [99ddbb74]
- @ai-sdk/ui-utils@0.0.14
- @ai-sdk/react@0.0.22
- @ai-sdk/solid@0.0.16
- @ai-sdk/svelte@0.0.17
- @ai-sdk/vue@0.0.17
## 3.2.24
### Patch Changes
- f041c056: feat (ai/core): add roundtrips property to generateText result
## 3.2.23
### Patch Changes
- a6cb2c8b: feat (ai/ui): add keepLastMessageOnError option to useChat
- Updated dependencies [a6cb2c8b]
- @ai-sdk/ui-utils@0.0.13
- @ai-sdk/svelte@0.0.16
- @ai-sdk/react@0.0.21
- @ai-sdk/solid@0.0.15
- @ai-sdk/vue@0.0.16
## 3.2.22
### Patch Changes
- 53fccf1c: fix (ai/core): report error on controller
- dd0d854e: feat (ai/vue): add useAssistant
- Updated dependencies [dd0d854e]
- @ai-sdk/vue@0.0.15
## 3.2.21
### Patch Changes
- 56bbc2a7: feat (ai/ui): set body and headers directly on options for handleSubmit and append
- Updated dependencies [56bbc2a7]
- @ai-sdk/ui-utils@0.0.12
- @ai-sdk/svelte@0.0.15
- @ai-sdk/react@0.0.20
- @ai-sdk/solid@0.0.14
- @ai-sdk/vue@0.0.14
## 3.2.20
### Patch Changes
- 671331b6: feat (core): add experimental OpenTelemetry support for generateText and streamText
## 3.2.19
### Patch Changes
- b7290943: chore (ai/core): rename TokenUsage type to CompletionTokenUsage
- b7290943: feat (ai/core): add token usage to embed and embedMany
- Updated dependencies [b7290943]
- @ai-sdk/provider@0.0.12
- @ai-sdk/provider-utils@1.0.2
- @ai-sdk/react@0.0.19
- @ai-sdk/svelte@0.0.14
- @ai-sdk/ui-utils@0.0.11
- @ai-sdk/solid@0.0.13
- @ai-sdk/vue@0.0.13
## 3.2.18
### Patch Changes
- Updated dependencies [70d18003]
- @ai-sdk/react@0.0.18
## 3.2.17
### Patch Changes
- 3db90c3d: allow empty handleSubmit submissions for useChat
- abb22602: feat (ai): verify that system messages have string content
- 5c1f0bd3: fix unclosed streamable value console message
- Updated dependencies [6a11cfaa]
- Updated dependencies [3db90c3d]
- Updated dependencies [d481729f]
- @ai-sdk/react@0.0.17
- @ai-sdk/svelte@0.0.13
- @ai-sdk/solid@0.0.12
- @ai-sdk/vue@0.0.12
- @ai-sdk/provider-utils@1.0.1
- @ai-sdk/ui-utils@0.0.10
## 3.2.16
### Patch Changes
- Updated dependencies [3f756a6b]
- @ai-sdk/react@0.0.16
## 3.2.15
### Patch Changes
- 6c99581e: fix (ai/react): stop() on useObject does not throw error and clears isLoading
- Updated dependencies [6c99581e]
- @ai-sdk/react@0.0.15
## 3.2.14
### Patch Changes
- Updated dependencies [9b50003d]
- Updated dependencies [1894f811]
- @ai-sdk/react@0.0.14
- @ai-sdk/ui-utils@0.0.9
- @ai-sdk/solid@0.0.11
- @ai-sdk/svelte@0.0.12
- @ai-sdk/vue@0.0.11
## 3.2.13
### Patch Changes
- d3100b9c: feat (ai/ui): support custom fetch function in useChat, useCompletion, useAssistant, useObject
- Updated dependencies [d3100b9c]
- @ai-sdk/ui-utils@0.0.8
- @ai-sdk/svelte@0.0.11
- @ai-sdk/react@0.0.13
- @ai-sdk/solid@0.0.10
- @ai-sdk/vue@0.0.10
## 3.2.12
### Patch Changes
- 5edc6110: feat (ai/core): add custom request header support
- Updated dependencies [5edc6110]
- Updated dependencies [5edc6110]
- Updated dependencies [5edc6110]
- @ai-sdk/provider@0.0.11
- @ai-sdk/provider-utils@1.0.0
- @ai-sdk/react@0.0.12
- @ai-sdk/svelte@0.0.10
- @ai-sdk/ui-utils@0.0.7
- @ai-sdk/solid@0.0.9
- @ai-sdk/vue@0.0.9
## 3.2.11
### Patch Changes
- c908f741: chore (ui/solid): update solidjs useChat and useCompletion to feature parity with React
- 827ef450: feat (ai/ui): improve error handling in useAssistant
- Updated dependencies [c908f741]
- Updated dependencies [827ef450]
- @ai-sdk/solid@0.0.8
- @ai-sdk/svelte@0.0.9
- @ai-sdk/react@0.0.11
## 3.2.10
### Patch Changes
- Updated dependencies [5b04204b]
- Updated dependencies [8f482903]
- @ai-sdk/react@0.0.10
## 3.2.9
### Patch Changes
- 82d9c8de: feat (ai/ui): make event in useAssistant submitMessage optional
- Updated dependencies [82d9c8de]
- Updated dependencies [321a7d0e]
- Updated dependencies [82d9c8de]
- @ai-sdk/svelte@0.0.8
- @ai-sdk/react@0.0.9
- @ai-sdk/vue@0.0.8
## 3.2.8
### Patch Changes
- 54bf4083: feat (ai/react): control request body in useChat
- Updated dependencies [54bf4083]
- @ai-sdk/ui-utils@0.0.6
- @ai-sdk/react@0.0.8
- @ai-sdk/solid@0.0.7
- @ai-sdk/svelte@0.0.7
- @ai-sdk/vue@0.0.7
## 3.2.7
### Patch Changes
- d42b8907: feat (ui): make event in handleSubmit optional
- Updated dependencies [d42b8907]
- @ai-sdk/svelte@0.0.6
- @ai-sdk/react@0.0.7
- @ai-sdk/solid@0.0.6
- @ai-sdk/vue@0.0.6
## 3.2.6
### Patch Changes
- 74e28222: fix (ai/rsc): "could not find InternalStreamableUIClient" bug
## 3.2.5
### Patch Changes
- 4d426d0c: fix (ai): split provider and model ids correctly in the provider registry
## 3.2.4
### Patch Changes
- Updated dependencies [3cb103bc]
- @ai-sdk/react@0.0.6
## 3.2.3
### Patch Changes
- 89b7552b: chore (ai): remove deprecation from ai/react imports, add experimental_useObject
- Updated dependencies [02f6a088]
- @ai-sdk/provider-utils@0.0.16
- @ai-sdk/react@0.0.5
- @ai-sdk/svelte@0.0.5
- @ai-sdk/ui-utils@0.0.5
- @ai-sdk/solid@0.0.5
- @ai-sdk/vue@0.0.5
## 3.2.2
### Patch Changes
- 0565cd72: feat (ai/core): add toJsonResponse to generateObject result.
## 3.2.1
### Patch Changes
- 008725ec: feat (ai): add textStream, toTextStreamResponse(), and pipeTextStreamToResponse() to streamObject
- 520fb2d5: feat (rsc): add streamUI onFinish callback
- Updated dependencies [008725ec]
- Updated dependencies [008725ec]
- @ai-sdk/react@0.0.4
- @ai-sdk/ui-utils@0.0.4
- @ai-sdk/solid@0.0.4
- @ai-sdk/svelte@0.0.4
- @ai-sdk/vue@0.0.4
## 3.2.0
### Minor Changes
- 85ef6d18: chore (ai): AI SDK 3.2 release
### Patch Changes
- b965dd2d: fix (core): pass settings correctly for generateObject and streamObject
## 3.1.37
### Patch Changes
- 85712895: chore (@ai-sdk/provider-utils): move test helper to provider utils
- Updated dependencies [85712895]
- Updated dependencies [85712895]
- @ai-sdk/provider-utils@0.0.15
- @ai-sdk/react@0.0.3
- @ai-sdk/svelte@0.0.3
- @ai-sdk/ui-utils@0.0.3
- @ai-sdk/solid@0.0.3
- @ai-sdk/vue@0.0.3
## 3.1.36
### Patch Changes
- 4728c37f: feat (core): add text embedding model support to provider registry
- 8c49166e: chore (core): rename experimental_createModelRegistry to experimental_createProviderRegistry
- Updated dependencies [7910ae84]
- @ai-sdk/provider-utils@0.0.14
- @ai-sdk/react@0.0.2
- @ai-sdk/svelte@0.0.2
- @ai-sdk/ui-utils@0.0.2
- @ai-sdk/solid@0.0.2
- @ai-sdk/vue@0.0.2
## 3.1.35
### Patch Changes
- 06123501: feat (core): support https and data url strings in image parts
## 3.1.34
### Patch Changes
- d25566ac: feat (core): add cosineSimilarity helper function
- 87a5d27e: feat (core): introduce InvalidMessageRoleError.
## 3.1.33
### Patch Changes
- 6fb14b5d: chore (streams): deprecate nanoid export.
- 05536768: feat (core): add experimental model registry
## 3.1.32
### Patch Changes
- 3cabf078: fix(ai/rsc): Refactor streamable UI internal implementation
## 3.1.31
### Patch Changes
- 85f209a4: chore: extracted ui library support into separate modules
- 85f209a4: removed (streams): experimental_StreamingReactResponse was removed. Please use AI SDK RSC instead.
- Updated dependencies [85f209a4]
- @ai-sdk/ui-utils@0.0.1
- @ai-sdk/svelte@0.0.1
- @ai-sdk/react@0.0.1
- @ai-sdk/solid@0.0.1
- @ai-sdk/vue@0.0.1
## 3.1.30
### Patch Changes
- fcf4323b: fix (core): filter out empty assistant text messages
## 3.1.29
### Patch Changes
- 28427d3e: feat (core): add streamObject onFinish callback
## 3.1.28
### Patch Changes
- 102ca22f: feat (core): add object promise to streamObject result
- Updated dependencies [102ca22f]
- @ai-sdk/provider@0.0.10
- @ai-sdk/provider-utils@0.0.13
## 3.1.27
### Patch Changes
- c9198d4d: feat (ui): send annotation and data fields in useChat when sendExtraMessageFields is true
- Updated dependencies [09295e2e]
- Updated dependencies [09295e2e]
- Updated dependencies [043a5de2]
- @ai-sdk/provider@0.0.9
- @ai-sdk/provider-utils@0.0.12
## 3.1.26
### Patch Changes
- 5ee44cae: feat (provider): langchain StringOutputParser support
## 3.1.25
### Patch Changes
- ff281126: fix(ai/rsc): Remove extra reconcilation of streamUI
## 3.1.24
### Patch Changes
- 93cae126: fix(ai/rsc): Fix unsafe {} type in application code for StreamableValue
- 08b5c509: feat (core): add tokenUsage to streamObject result
## 3.1.23
### Patch Changes
- c03cafe6: chore (core, ui): rename maxAutomaticRoundtrips to maxToolRoundtrips
## 3.1.22
### Patch Changes
- 14bb8694: chore (ui): move maxAutomaticRoundtrips and addToolResult out of experimental
## 3.1.21
### Patch Changes
- 213f2411: fix (core,streams): support ResponseInit variants
- 09698bca: chore (streams): deprecate streaming helpers that have a provider replacement
## 3.1.20
### Patch Changes
- 0e1da476: feat (core): add maxAutomaticRoundtrips setting to generateText
## 3.1.19
### Patch Changes
- 9882d24b: fix (ui/svelte): send data to server
- 131bbd3e: fix (ui): remove console.log statements
## 3.1.18
### Patch Changes
- f9dee8ac: fix(ai/rsc): Fix types for createStreamableValue and createStreamableUI
- 1c0ebf8e: feat (core): add responseMessages to generateText result
## 3.1.17
### Patch Changes
- 92b993b7: ai/rsc: improve getAIState and getMutableAIState types
- 7de628e9: chore (ui): deprecate old function/tool call handling
- 7de628e9: feat (ui): add onToolCall handler to useChat
## 3.1.16
### Patch Changes
- f39c0dd2: feat (core, rsc): add toolChoice setting
- Updated dependencies [f39c0dd2]
- @ai-sdk/provider@0.0.8
- @ai-sdk/provider-utils@0.0.11
## 3.1.15
### Patch Changes
- 8e780288: feat (ai/core): add onFinish callback to streamText
- 8e780288: feat (ai/core): add text, toolCalls, and toolResults promises to StreamTextResult (matching the generateText result API with async methods)
- Updated dependencies [8e780288]
- @ai-sdk/provider@0.0.7
- @ai-sdk/provider-utils@0.0.10
## 3.1.14
### Patch Changes
- 6109c6a: feat (ai/react): add experimental_maxAutomaticRoundtrips to useChat
## 3.1.13
### Patch Changes
- 60117c9: dependencies (ai/ui): add React 18.3 and 19 support (peer dependency)
- Updated dependencies [6a50ac4]
- Updated dependencies [6a50ac4]
- @ai-sdk/provider@0.0.6
- @ai-sdk/provider-utils@0.0.9
## 3.1.12
### Patch Changes
- ae05fb7: feat (ai/streams): add StreamData support to streamToResponse
## 3.1.11
### Patch Changes
- a085d42: fix (ai/ui): decouple StreamData chunks from LLM stream
## 3.1.10
### Patch Changes
- 3a21030: feat (ai/core): add embedMany function
## 3.1.9
### Patch Changes
- 18a9655: feat (ai/svelte): add useAssistant
## 3.1.8
### Patch Changes
- 0f6bc4e: feat (ai/core): add embed function
- Updated dependencies [0f6bc4e]
- @ai-sdk/provider@0.0.5
- @ai-sdk/provider-utils@0.0.8
## 3.1.7
### Patch Changes
- f617b97: feat (ai): support client/server tool calls with useChat and streamText
## 3.1.6
### Patch Changes
- 2e78acb: Deprecate StreamingReactResponse (use AI SDK RSC instead).
- 8439884: ai/rsc: make RSC streamable utils chainable
- 325ca55: feat (ai/core): improve image content part error message
- Updated dependencies [325ca55]
- @ai-sdk/provider@0.0.4
- @ai-sdk/provider-utils@0.0.7
## 3.1.5
### Patch Changes
- 5b01c13: feat (ai/core): add system message support in messages list
## 3.1.4
### Patch Changes
- ceb44bc: feat (ai/ui): add stop() helper to useAssistant (important: AssistantResponse now requires OpenAI SDK 4.42+)
- 37c9d4c: feat (ai/streams): add LangChainAdapter.toAIStream()
## 3.1.3
### Patch Changes
- 970a099: fix (ai/core): streamObject fixes partial json with empty objects correctly
- 1ac2390: feat (ai/core): add usage and finishReason to streamText result.
- Updated dependencies [276f22b]
- @ai-sdk/provider-utils@0.0.6
## 3.1.2
### Patch Changes
- d1b1880: fix (ai/core): allow reading streams in streamText result multiple times
## 3.1.1
### Patch Changes
- 0f77132: ai/rsc: remove experimental\_ from streamUI
## 3.1.0
### Minor Changes
- 73356a9: Move AI Core functions out of experimental (streamText, generateText, streamObject, generateObject).
## 3.0.35
### Patch Changes
- 41d5736: ai/core: re-expose language model types.
- b4c68ec: ai/rsc: ReadableStream as provider for createStreamableValue; add .append() method
- Updated dependencies [41d5736]
- @ai-sdk/provider@0.0.3
- @ai-sdk/provider-utils@0.0.5
## 3.0.34
### Patch Changes
- b9a831e: ai/rsc: add experimental_streamUI()
## 3.0.33
### Patch Changes
- 56ef84a: ai/core: fix abort handling in transformation stream
- Updated dependencies [56ef84a]
- @ai-sdk/provider-utils@0.0.4
## 3.0.32
### Patch Changes
- 0e0d2af: ai/core: add pipeTextStreamToResponse helper to streamText.
## 3.0.31
### Patch Changes
- 74c63b1: ai/core: add toAIStreamResponse() helper to streamText.
## 3.0.30
### Patch Changes
- e7e5898: use-assistant: fix missing message content
## 3.0.29
### Patch Changes
- 22a737e: Fix: mark useAssistant as in progress for append/submitMessage.
## 3.0.28
### Patch Changes
- d6431ae: ai/core: add logprobs support (thanks @SamStenner for the contribution)
- 25f3350: ai/core: add support for getting raw response headers.
- Updated dependencies [d6431ae]
- Updated dependencies [25f3350]
- @ai-sdk/provider@0.0.2
- @ai-sdk/provider-utils@0.0.3
## 3.0.27
### Patch Changes
- eb150a6: ai/core: remove scaling of setting values (breaking change). If you were using the temperature, frequency penalty, or presence penalty settings, you need to update the providers and adjust the setting values.
- Updated dependencies [eb150a6]
- @ai-sdk/provider-utils@0.0.2
- @ai-sdk/provider@0.0.1
## 3.0.26
### Patch Changes
- f90f6a1: ai/core: add pipeAIStreamToResponse() to streamText result.
## 3.0.25
### Patch Changes
- 1e84d6d: Fix: remove mistral lib type dependency.
- 9c2a049: Add append() helper to useAssistant.
## 3.0.24
### Patch Changes
- e94fb32: feat(ai/rsc): Make `onSetAIState` and `onGetUIState` stable
## 3.0.23
### Patch Changes
- 66b5892: Add streamMode parameter to useChat and useCompletion.
- Updated dependencies [7b8791d]
- @ai-sdk/provider-utils@0.0.1
## 3.0.22
### Patch Changes
- d544886: Breaking change: extract experimental AI core provider packages. They can now be imported with e.g. import { openai } from '@ai-sdk/openai' after adding them to a project.
- ea6b0e1: Expose formatStreamPart, parseStreamPart, and readDataStream helpers.
## 3.0.21
### Patch Changes
- 87d3db5: Extracted @ai-sdk/provider package
- 8c40f8c: ai/core: Fix openai provider streamObject for gpt-4-turbo
- 5cd29bd: ai/core: add toTextStreamResponse() method to streamText result
## 3.0.20
### Patch Changes
- f42bbb5: Remove experimental from useAssistant and AssistantResponse.
- 149fe26: Deprecate <Tokens/>
- 2eb4b55: Remove experimental\_ prefix from StreamData.
- e45fa96: Add stream support for Bedrock/Cohere.
- a6b2500: Deprecated the `experimental_streamData: true` setting from AIStreamCallbacksAndOptions. You can delete occurrences in your code. The stream data protocol is now used by default.
## 3.0.19
### Patch Changes
- 4f4c7f5: ai/core: Anthropic tool call support
## 3.0.18
### Patch Changes
- 63d587e: Add Anthropic provider for ai/core functions (no tool calling).
- 63d587e: Add automatic mime type detection for images in ai/core prompts.
## 3.0.17
### Patch Changes
- 2b991c4: Add Google Generative AI provider for ai/core functions.
## 3.0.16
### Patch Changes
- a54ea77: feat(ai/rsc): add `useStreamableValue`
## 3.0.15
### Patch Changes
- 4aed2a5: Add JSDoc comments for ai/core functions.
- cf8d12f: Export experimental language model specification under `ai/spec`.
## 3.0.14
### Patch Changes
- 8088de8: fix(ai/rsc): improve typings for `StreamableValue`
- 20007b9: feat(ai/rsc): support string diff and patch in streamable value
- 6039460: Support Bedrock Anthropic Stream for Messages API.
- e83bfe3: Added experimental ai/core functions (streamText, generateText, streamObject, generateObject). Add OpenAI and Mistral language model providers.
## 3.0.13
### Patch Changes
- 026d061: Expose setMessages in useAssistant hook
- 42209be: AssistantResponse: specify forwardStream return type.
## 3.0.12
### Patch Changes
- b99b008: fix(ai/rsc): avoid appending boundary if the same reference was passed
## 3.0.11
### Patch Changes
- ce009e2: Added OpenAI assistants streaming.
- 3f9bf3e: Updates types to OpenAI SDK 4.29.0
## 3.0.10
### Patch Changes
- 33d261a: fix(ai/rsc): Fix .append() behavior
## 3.0.9
### Patch Changes
- 81ca3d6: fix(ai/rsc): improve .done() argument type
## 3.0.8
### Patch Changes
- a94aab2: ai/rsc: optimize streamable value stream size
## 3.0.7
### Patch Changes
- 9a9ae73: feat(ai/rsc): readStreamableValue
## 3.0.6
### Patch Changes
- 1355ad0: Fix: experimental_onToolCall is called with parsed tool args
- 9348f06: ai/rsc: improve dev error and warnings by trying to detect hanging streams
- 8be9404: fix type resolution
## 3.0.5
### Patch Changes
- a973f1e: Support Anthropic SDK v0.15.0
- e25f3ca: type improvements
## 3.0.4
### Patch Changes
- 7962862: fix `useActions` type inference
- aab5324: Revert "fix(render): parse the args based on the zod schema"
- fe55612: Bump OpenAI dependency to 4.28.4; fix type error in render
## 3.0.3
### Patch Changes
- 4d816ca: fix(render): parse the args based on the zod schema
- d158a47: fix potential race conditions
## 3.0.2
### Patch Changes
- 73bd06e: fix(useActions): return typed object
## 3.0.1
### Patch Changes
- ac20a25: ai/rsc: fix text response and async generator
- b88778f: Added onText callback for text tokens.
## 3.0.0
### Major Changes
- 51054a9: add ai/rsc
## 2.2.37
### Patch Changes
- a6b5764: Add support for Mistral's JavaScript SDK
## 2.2.36
### Patch Changes
- 141f0ce: Fix: onFinal callback is invoked with text from onToolCall when onToolCall returns string
## 2.2.35
### Patch Changes
- b717dad: Adding Inkeep as a stream provider
## 2.2.34
### Patch Changes
- 2c8ffdb: cohere-stream: support AsyncIterable
- ed1e278: Message annotations handling for all Message types
## 2.2.33
### Patch Changes
- 8542ae7: react/use-assistant: add onError handler
- 97039ff: OpenAIStream: Add support for the Azure OpenAI client library
## 2.2.32
### Patch Changes
- 7851fa0: StreamData: add `annotations` and `appendMessageAnnotation` support
## 2.2.31
### Patch Changes
- 9b89c4d: react/use-assistant: Expose setInput
- 75751c9: ai/react: Add experimental_onToolCall to useChat.
## 2.2.30
### Patch Changes
- ac503e0: ai/solid: add chat request options to useChat
- b78a73e: Add GoogleGenerativeAIStream for Gemini support
- 5220336: ai/svelte: Add experimental_onToolCall to useChat.
- ef99062: Add support for the Anthropic message API
- 5220336: Add experimental_onToolCall to OpenAIStream.
- ac503e0: ai/vue: add chat request options to useChat
## 2.2.29
### Patch Changes
- 5a9ae2e: ai/prompt: add `experimental_buildOpenAIMessages` to validate and cast AI SDK messages to OpenAI messages
## 2.2.28
### Patch Changes
- 07a679c: Add data message support to useAssistant & assistantResponse.
- fbae595: ai/react: `api` functions are no longer used as a cache key in `useChat`
## 2.2.27
### Patch Changes
- 0fd1205: ai/vue: Add complex response parsing and StreamData support to useCompletion
- a7dc746: experimental_useAssistant: Expose extra fetch options
- 3dcf01e: ai/react Add data support to useCompletion
- 0c3b338: ai/svelte: Add complex response parsing and StreamData support to useCompletion
- 8284777: ai/solid: Add complex response parsing and StreamData support to useCompletion
## 2.2.26
### Patch Changes
- df1ad33: ai/vue: Add complex response parsing and StreamData support to useChat
- 3ff8a56: Add `generateId` to use-chat params to allow overriding message ID generation
- 6c2a49c: ai/react experimental_useAssistant() submit can be called without an event
- 8b4f7d1: ai/react: Add complex response parsing and StreamData support to useCompletion
## 2.2.25
### Patch Changes
- 1e61c69: chore: specify the minimum react version to 18
- 6aec2d2: Expose threadId in useAssistant
- c2369df: Add AWS Bedrock support
- 223fde3: ai/svelte: Add complex response parsing and StreamData support to useChat
## 2.2.24
### Patch Changes
- 69ca8f5: ai/react: add experimental_useAssistant hook and experimental_AssistantResponse
- 3e2299e: experimental_StreamData/StreamingReactResponse: optimize parsing, improve types
- 70bd2ac: ai/solid: add experimental_StreamData support to useChat
## 2.2.23
### Patch Changes
- 5a04321: add StreamData support to StreamingReactResponse, add client-side data API to react/use-chat
## 2.2.22
### Patch Changes
- 4529831: ai/react: Do not store initialMessages in useState
- db5378c: experimental_StreamData: fix data type to be JSONValue
## 2.2.21
### Patch Changes
- 2c8d4bd: Support openai@4.16.0 and later
## 2.2.20
### Patch Changes
- 424d5ee: experimental_StreamData: fix trailing newline parsing bug in decoder
- c364c6a: cohere: fix closing cohere stream, avoids response from hanging
## 2.2.19
### Patch Changes
- 699552d: add experimental_StreamingReactResponse
## 2.2.18
### Patch Changes
- 0bd27f6: react/use-chat: allow client-side handling of function call without following response
## 2.2.17
### Patch Changes
- 5ed581d: Use interface instead of type for Message to allow declaration merging
- 9adec1e: vue and solid: fix including `function_call` and `name` fields in subsequent requests
## 2.2.16
### Patch Changes
- e569688: Fix for #637, resync interfaces
## 2.2.15
### Patch Changes
- c5d1857: fix: return complete response in onFinish when onCompletion isn't passed
- c5d1857: replicate-stream: fix types for replicate@0.20.0+
## 2.2.14
### Patch Changes
- 6229d6b: openai: fix OpenAIStream types with openai@4.11+
## 2.2.13
### Patch Changes
- a4a997f: all providers: reset error message on (re)submission
## 2.2.12
### Patch Changes
- cb181b4: ai/vue: wrap body with unref to support reactivity
## 2.2.11
### Patch Changes
- 2470658: ai/react: fix: handle partial chunks in react getStreamedResponse when using experimental_StreamData
## 2.2.10
### Patch Changes
- 8a2cbaf: vue/use-completion: fix: don't send network request for loading state"
- bbf4403: langchain-stream: return langchain `writer` from LangChainStream
## 2.2.9
### Patch Changes
- 3fc2b32: ai/vue: fix: make body parameter reactive
## 2.2.8
### Patch Changes
- 26bf998: ai/react: make reload/complete/append functions stable via useCallback
## 2.2.7
### Patch Changes
- 2f97630: react/use-chat: fix aborting clientside function calls too early
- 1157340: fix: infinite loop for experimental stream data (#484)
## 2.2.6
### Patch Changes
- e5bf68d: react/use-chat: fix experimental functions returning proper function messages
Closes #478
## 2.2.5
### Patch Changes
- e5bf68d: react/use-chat: fix experimental functions returning proper function messages
Closes #478
## 2.2.4
### Patch Changes
- 7b389a7: fix: improve safety for type check in openai-stream
## 2.2.3
### Patch Changes
- 867a3f9: Fix client-side function calling (#467, #469)
add Completion type from the `openai` SDK to openai-stream (#472)
## 2.2.2
### Patch Changes
- 84e0cc8: Add experimental_StreamData and new opt-in wire protocol to enable streaming additional data. See https://github.com/vercel/ai/pull/425.
Changes `onCompletion` back to run every completion, including recursive function calls. Adds an `onFinish` callback that runs once everything has streamed.
If you're using experimental function handlers on the server _and_ caching via `onCompletion`,
you may want to adjust your caching code to account for recursive calls so the same key isn't used.
```
let depth = 0
const stream = OpenAIStream(response, {
async onCompletion(completion) {
depth++
await kv.set(key + '_' + depth, completion)
await kv.expire(key + '_' + depth, 60 * 60)
}
})
```
## 2.2.1
### Patch Changes
- 04084a8: openai-stream: fix experimental_onFunctionCall types for OpenAI SDK v4
## 2.2.0
### Minor Changes
- dca1ed9: Update packages and examples to use OpenAI SDK v4
## 2.1.34
### Patch Changes
- c2917d3: Add support for the Anthropic SDK, newer Anthropic API versions, and improve Anthropic error handling
## 2.1.33
### Patch Changes
- 4ef8015: Prevent `isLoading` in vue integration from triggering extraneous network requests
## 2.1.32
### Patch Changes
- 5f91427: ai/svelte: fix isLoading return value
## 2.1.31
### Patch Changes
- ab2b973: fix pnpm-lock.yaml
## 2.1.30
### Patch Changes
- 4df2a49: Fix termination of ReplicateStream by removing the terminating `{}`from output
## 2.1.29
### Patch Changes
- 3929a41: Add ReplicateStream helper
## 2.1.28
### Patch Changes
- 9012e17: react/svelte/vue: fix making unnecessary SWR request to API endpoint
## 2.1.27
### Patch Changes
- 3d29799: React/Svelte/Vue: keep isLoading in sync between hooks with the same ID.
React: don't throw error when submitting
## 2.1.26
### Patch Changes
- f50d9ef: Add experimental_buildLlama2Prompt helper for Hugging Face
## 2.1.25
### Patch Changes
- 877c16f: ai/react: don't throw error if onError is passed
## 2.1.24
### Patch Changes
- f3f5866: Adds SolidJS support and SolidStart example
## 2.1.23
### Patch Changes
- 0ebc2f0: streams/openai-stream: don't call onStart/onCompletion when recursing
## 2.1.22
### Patch Changes
- 9320e95: Add (experimental) prompt construction helpers for StarChat and OpenAssistant
- e3a7ec8: Support <|end|> token for StarChat beta in huggingface-stream
## 2.1.21
### Patch Changes
- 561a49a: Providing a function to `function_call` request parameter of the OpenAI Chat Completions API no longer breaks OpenAI function stream parsing.
## 2.1.20
### Patch Changes
- e361114: OpenAI functions: allow returning string in callback
## 2.1.19
### Patch Changes
- e4281ca: Add experimental server-side OpenAI function handling
## 2.1.18
### Patch Changes
- 6648b21: Add experimental client side OpenAI function calling to Svelte bindings
- e5b983f: feat(streams): add http error handling for openai
## 2.1.17
### Patch Changes
- 3ed65bf: Remove dependency on node crypto API
## 2.1.16
### Patch Changes
- 8bfb43d: Fix svelte peer dependency version
## 2.1.15
### Patch Changes
- 4a2b978: Update cohere stream and add docs
## 2.1.14
### Patch Changes
- 3164adb: Fix regression with generated ids
## 2.1.13
### Patch Changes
- fd82961: Use rfc4122 IDs when generating chat/completion IDs
## 2.1.12
### Patch Changes
- b7b93e5: Add <Tokens> RSC to ai/react
## 2.1.11
### Patch Changes
- 8bf637a: Fix langchain handlers so that they now are correctly invoked and update examples and docs to show correct usage (passing the handlers to `llm.call` and not the model itself).
## 2.1.10
### Patch Changes
- a7b3d0e: Experimental support for OpenAI function calling
## 2.1.9
### Patch Changes
- 9cdf968: core/react: add Tokens react server component
## 2.1.8
### Patch Changes
- 44d9879: Support extra request options in chat and completion hooks
## 2.1.7
### Patch Changes
- bde3898: Allow an async onResponse callback in useChat/useCompletion
## 2.1.6
### Patch Changes
- 23f0899: Set stream: true when decoding streamed chunks
## 2.1.5
### Patch Changes
- 89938b0: Provider direct callback handlers in LangChain now that `CallbackManager` is deprecated.
## 2.1.4
### Patch Changes
- c16d650: Improve type saftey for AIStream. Added JSDoc comments.
## 2.1.3
### Patch Changes
- a9591fe: Add `createdAt` on `user` input message in `useChat` (it was already present in `assistant` messages)
## 2.1.2
### Patch Changes
- f37d4ec: fix bundling
## 2.1.1
### Patch Changes
- 9fdb51a: fix: add better typing for store within svelte implementation (#104)
## 2.1.0
### Minor Changes
- 71f9c51: This adds Vue support for `ai` via the `ai/vue` subpath export. Vue composables `useChat` and `useCompletion` are provided.
### Patch Changes
- ad54c79: add tests
## 2.0.1
### Patch Changes
- be90740: - Switches `LangChainStream` helper callback `handler` to return use `handleChainEnd` instead of `handleLLMEnd` so as to work with sequential chains
## 2.0.0
### Major Changes
- 095de43: New package name!
## 0.0.14
### Patch Changes
- c6586a2: Add onError callback, include response text in error if response is not okay
## 0.0.13
### Patch Changes
- c1f4a91: Throw error when provided AI response isn't valid
## 0.0.12
### Patch Changes
- ea4e66a: improve API types
## 0.0.11
### Patch Changes
- a6bc35c: fix package exports for react and svelte subpackages
## 0.0.10
### Patch Changes
- 56f9537: add svelte apis
## 0.0.9
### Patch Changes
- 78477d3: - Create `/react` sub-package.
- Create `import { useChat, useCompletion } from 'ai/react'` and mark React as an optional peer dependency so we can add more framework support in the future.
- Also renamed `set` to `setMessages` and `setCompletion` to unify the API naming as we have `setInput` too.
- Added an `sendExtraMessageFields` field to `useChat` that defaults to `false`, to prevent OpenAI errors when `id` is not filtered out.
- c4c1be3: useCompletion.handleSubmit does not clear the input anymore
- 7de2185: create /react export
## 0.0.8
### Patch Changes
- fc83e95: Implement new start-of-stream newline trimming
- 2c6fa04: Optimize callbacks TransformStream to be more memory efficient when `onCompletion` is not specified
## 0.0.7
### Patch Changes
- fdfef52: - Splits the `EventSource` parser into a reusable helper
- Uses a `TransformStream` for this, so the stream respects back-pressure
- Splits the "forking" stream for callbacks into a reusable helper
- Changes the signature for `customParser` to avoid Stringify -> Encode -> Decode -> Parse round trip
- Uses ?.() optional call syntax for callbacks
- Uses string.includes to perform newline checking
- Handles the `null` `res.body` case
- Fixes Anthropic's streaming responses
- Anthropic returns cumulative responses, not deltas like OpenAI
- https://github.com/hwchase17/langchain/blob/3af36943/langchain/llms/anthropic.py#L190-L193
## 0.0.6
### Patch Changes
- d70a9e7: Add streamToResponse
- 47b85b2: Improve abortController and callbacks of `useChat`
- 6f7b43a: Export `UseCompletionHelpers` as a TypeScript type alias
## 0.0.5
### Patch Changes
- 4405a8a: fix duplicated `'use client'` directives
## 0.0.4
### Patch Changes
- b869104: Added `LangChainStream`, `useCompletion`, and `useChat`
## 0.0.3
### Patch Changes
- 677d222: add useCompletion
## 0.0.2
### Patch Changes
- af400e2: Fix release script
## 0.0.1
### Patch Changes
- b7e227d: Add `useChat` hook
## 0.0.2
### Patch Changes
- 9a8a845: Testing out release
