VoiceLabs

Transcribe audio

transcribe

Transcribe an audio file to text using VoiceLabs' Whisper. Give the audio exactly one way: audioUrl, a public https:// URL serving the file itself (wav, mp3, m4a, aac, ogg, flac, aiff or webm; up to 10 MiB), or audioBase64, the file's real bytes (≤10 MiB of base64 text). A payload that is not real audio — under 1 KiB, a bare header, or made-up placeholder bytes — is refused without being transcribed. If the user has an audio file you cannot read the bytes of, ask them for a public link, or to upload it on the Captures page of the VoiceLabs studio (https://app.voicelabs.now/studio). Returns the transcript and the created capture id.

Input Schema

Name	Required	Description
`audioUrl`	No	A public https:// URL that serves the audio file itself (wav, mp3, m4a, aac, ogg, flac, aiff or webm; no redirects, up to 10 MiB). Give exactly one of audioBase64 or audioUrl.
`language`	No	ISO language code (auto-detected when omitted).
`audioBase64`	No	The audio file's real bytes as base64 (≤10 MiB of base64 text; at least 1 KiB of audio). Only use this when you actually hold the file's bytes — a made-up or placeholder payload is refused. Give exactly one of audioBase64 or audioUrl.

Output Schema

Name	Required	Description	Default
`language`	Yes
`captureId`	Yes
`durationMs`	Yes
`transcript`	Yes

Schema Changelog

Changes observed during successful MCP inspections.

Changed4 schema fields changedOct 26, 2026

removedOutput schema / properties / durationMs / anyOf

Removed value: -[
-  {
-    "type": "number"
-  },
-  {
-    "type": "null"
-  }
-]

addedOutput schema / properties / durationMs / type
```
Added value: +[
+  "number",
+  "null"
+]
```

removedOutput schema / properties / language / anyOf

Removed value: -[
-  {
-    "type": "string"
-  },
-  {
-    "type": "null"
-  }
-]

addedOutput schema / properties / language / type
```
Added value: +[
+  "string",
+  "null"
+]
```

First observedSep 28, 2026

TDQS

A4.3/5.0

Behavior4/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

Annotations declare readOnlyHint=false but don't explain why; the description supplies the reason by disclosing that a capture is created and its id returned. It also discloses non-obvious rejection behavior ('A payload that is not real audio — under 1 KiB, a bare header, or made-up placeholder bytes — is refused without being transcribed'), which saves an agent from wasting a call. Missing: whether the call blocks for the duration of transcription, and any auth/quota context.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness4/5

Is the description appropriately sized, front-loaded, and free of redundancy?

Front-loaded with the action, then the input contract, then the fallback. Every sentence carries instruction value, but the inline format list (wav, mp3, m4a, ...) and the 10 MiB ceiling partially duplicate the schema and make the block longer than necessary.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness4/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

For a 3-parameter, zero-required tool with an output schema, the description covers the hard parts: mutually exclusive inputs, size/format constraints, validation rejection, and the human-in-the-loop fallback. The remaining gap is execution semantics — synchronous vs. long-running transcription — which matters for an audio tool but is nowhere stated.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters4/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema coverage is 100%, so the baseline is 3, but the description adds meaning beyond the schema: it warns that the URL must serve the file itself, that the base64 must be the file's genuine bytes, and gives the 1 KiB lower bound for plausible audio plus the user-facing workaround. The `language` parameter is left unmentioned in the description, though its auto-detect behavior is fully documented in the schema.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

States a specific verb and resource ('Transcribe an audio file to text using VoiceLabs' Whisper') and names the engine, which is enough to distinguish it from the TTS sibling `speak` and the retrieval siblings (`list_captures`, `get_generation`). It also states the outputs produced (transcript plus capture id), so an agent knows the effect of the call.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines4/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

Gives explicit input-selection rules ('Give the audio exactly one way: audioUrl ... or audioBase64') and a concrete fallback workflow when bytes are unreadable ('ask them for a public link, or to upload it on the Captures page'). It stops short of naming when to prefer this over a sibling or when not to call it at all, but the operational context is clear.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

All VoiceLabs tools Inspect

Try in Browser

One endpoint for every MCP toolRayrun runs every MCP server behind one URL and holds the credentials. Each call is decided per client, tool, and argument, then recorded.ray.runAd

Glama MCP Gateway

Add one secure layer between your agents and this server.

Resources

Need Help?