TokenLab MCP Server
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| TOKENLAB_API_KEY | No | Optional API key required for inference helper tools | |
| TOKENLAB_API_BASE | No | Optional base URL for the TokenLab API | https://api.tokenlab.sh |
Instructions
Guidance the server publishes about itself, which clients place ahead of the tool catalog so the model reads it before choosing anything.
This server publishes no instructions, or was last inspected before Glama recorded them.
Capabilities
Features and capabilities supported by this server
Protocol revision2025-11-25
| Capability | Details |
|---|---|
| tools | {} |
| prompts | {
"listChanged": true
} |
| resources | {
"listChanged": true
} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| cancel_taskA | Cancels a queued asynchronous task when cancellation is supported for the selected task. Cancellation currently supports queued Seedance video tasks ( |
| create_3d_modelB | Creates a 3D model generation task using Tripo3D. Returns a task ID for polling. |
| create_anthropic_messageA | Creates a model response using Anthropic's native request format. Supports Claude models with vision and streaming. Authentication: Use |
| create_chat_completionA | Create an OpenAI-compatible chat completion. For inline image_url data URLs, declare the byte-accurate image MIME type. The MCP boundary corrects recognized PNG, JPEG, WebP, and GIF payloads declared as application/octet-stream and rejects unrecognized generic binary image payloads before sending a billable request. |
| create_embeddingC | Creates an embedding vector representing the input text. |
| create_gemini_contentA | Generates content using the Gemini GenerateContent request shape. The model details must advertise Gemini requests and the model must be available. ProtoJSON lowerCamelCase and original proto snake_case field names are preserved. Unknown fields are accepted on a best-effort basis; model-specific support is not guaranteed. Authentication: Use |
| create_imageC | Creates an image given a prompt. |
| create_image_fileD | Creates an image given a prompt. |
| create_multimodal_embeddingB | Creates embeddings for multimodal input items. Text input is generally available; image input may require feature enablement. |
| create_musicC | Creates a music generation task using Suno. Returns a task ID for polling. |
| create_responseA | Creates a response using the OpenAI Responses API request shape. The model details must advertise the Responses request format and the model must be available. Unknown fields are accepted on a best-effort basis; model-specific support is not guaranteed. |
| create_speechC | Generates audio from the input text (Text-to-Speech). |
| create_videoA | Creates an asynchronous video generation task. The response returns a canonical task ID and usually a preferred |
| delete_fileB | Deletes a batch file or a bound image/reference file. When the request includes anthropic-beta: files-api-2025-04-14, the route uses Anthropic Files API mode for the bound workspace. |
| edit_imageB | Edits an image using multipart image uploads, JSON image URLs, or the official JSON images array for supported GPT Image models. |
| edit_image_fileC | Edits an image using multipart image uploads, JSON image URLs, or the official JSON images array for supported GPT Image models. |
| evaluate_decisionsA | Evaluate typed Noul probability, Choice, and Score questions against shared state using the native System One endpoint. Discover a decision model with list_models/get_model first. Returns synchronous structured decisions, not chat text. A decision does not authorize an external action. |
| get_modelA | Retrieves current model details, including supported operations, request parameters, request endpoints, selector limits, and pricing when available. Non-chat integrations should read these current model-specific request details before creating a request. |
| get_model_pricingA | Retrieves pricing-only detail for one model. Use this endpoint for price explanation, not for non-chat request construction. |
| get_pricingB | Returns the public pricing surface for active models, with optional provider and tag filters. |
| get_task_statusA | Retrieves the status and result of an asynchronous generation task. Prefer the |
| list_filesD | List files |
| list_modelsA | Lists the currently available models. Use |
| rerank_documentsC | Reranks documents by relevance to a query using semantic similarity. |
| retrieve_fileD | Retrieve file |
| retrieve_file_contentD | Retrieve file content |
| transcribe_audioC | Transcribes audio into text (Speech-to-Text). |
| translate_audioC | Transcribes audio and translates the result to English. |
| translate_textC | Translates text into a target language using the current translation request format. |
| upload_fileA | Uploads a TokenLab/OpenAI-compatible batch or image file. With anthropic-beta: files-api-2025-04-14, uploads an Anthropic Files resource without a purpose field and returns Anthropic file metadata. |
| compare_modelsA | Compare public TokenLab model details and pricing for several model IDs. |
| get_api_overviewA | Fetch TokenLab's agent-readable API overview. |
Prompts
Interactive templates invoked by user choice
| Name | Description |
|---|---|
| choose_tokenlab_model | Guide an agent through live model discovery and cost-aware comparison. |
| build_tokenlab_request | Guide an agent to produce a request that preserves the selected native endpoint contract. |
Resources
Contextual data attached and managed by the client
| Name | Description |
|---|---|
| tokenlab-api-overview | Live agent-readable overview of TokenLab endpoints, models, and recovery guidance. |
| tokenlab-openapi-contract | OpenAPI snapshot used to generate this MCP package version. |
| tokenlab-mcp-public-contract | Machine-readable package identity, profiles, tools, resources, and prompts. |
TDQS
Scored across 32 tools
Several tools are near-duplicates or have overlapping purposes: create_image and create_image_file both say 'Creates an image given a prompt,' and edit_image and edit_image_file share identical descriptions. Provider-specific response tools (create_chat_completion, create_anthropic_message, create_gemini_content, create_response) and model/pricing discovery tools (get_pricing, get_model_pricing, compare_models, list_models, get_model) force agents to guess which one fits a given request.
Every tool uses lowercase snake_case, and the names consistently follow a verb_noun pattern (create_, get_, list_, retrieve_, delete_, cancel_, edit_, compare_, etc.). There are no mixed conventions such as camelCase or vague standalone verbs.
32 tools is above the typical well-scoped range and crosses the 'too many' threshold. The broad TokenLab platform explains some breadth, but redundant image/file aliases and multiple provider-specific generation tools inflate the surface beyond what an agent can easily navigate.
The surface covers a wide AI platform: chat completions, embeddings, image generation/editing, speech, transcription, translation, video, music, 3D, files, models, pricing, async task status, and decision evaluation. Minor gaps exist, notably cancel_task only supports queued Seedance video tasks and there are no update operations for file/model resources, but core workflows are broadly covered.