Mistral MCP — Document Extraction

Overview Schema Source Related Servers Discussions Admin

Voxtral speech-to-text

voxtral_transcribe

Read-onlyIdempotent

Transcribe audio to text from a public URL or uploaded file, returning plain text, detected language, and optional speaker segments and timestamps.

Instructions

Transcribe an audio file to text using Mistral Voxtral.

Accepted models:

voxtral-mini-latest
voxtral-small-latest

Audio source is one of:

{ type: "file_url", fileUrl: "https://..." } (public URL)
{ type: "file", fileId: "" }

Options:

language: ISO-639-1 hint (e.g. 'fr', 'en'). Boosts accuracy when known.
temperature: sampling temperature.
diarize: return per-speaker segments (default false).
timestampGranularities: ['segment'] to return per-segment timestamps.
contextBias: list of phrases/terms that should bias the decoder.

Returns plain text, detected language, optional segments[], and token usage.

Input Schema

TableJSON Schema

Name	Required	Description
`audio`	Yes	Audio to transcribe, supplied as a public URL or an uploaded file ID.
`model`	No	STT model. Default: voxtral-mini-latest.
`diarize`	No	Identify speakers in the returned transcription segments. Defaults to false.
`language`	No	ISO-639-1 language hint (e.g. 'fr', 'en').
`contextBias`	No	Words or phrases to favor when decoding the audio.
`temperature`	No	Sampling temperature for transcription.
`timestampGranularities`	No	Only 'segment' is currently supported.

Output Schema

TableJSON Schema

Name	Required	Description	Default
`text`	Yes
`model`	Yes
`usage`	No
`language`	Yes
`segments`	No

Schema Changelog

Changes observed during successful MCP inspections.

Changed13 schema fields changedv0.8.3Sep 28, 2026

changedInput schema / $schema

Previous value: -"http://json-schema.org/draft-07/schema#"New value: +"https://json-schema.org/draft/2020-12/schema"

removedInput schema / additionalProperties
```
Removed value: -false
```

changedInput schema / properties / audio / anyOf

Previous value: -[
-  {
-    "additionalProperties": false,
-    "properties": {
-      "fileUrl": {
-        "description": "HTTPS URL to an audio file (mp3/wav/flac/ogg/webm/m4a).",
-        "type": "string"
-      },
-      "type": {
-        "const": "file_url",
-        "type": "string"
-      }
-    },
-    "required": [
-      "type",
-      "fileUrl"
-    ],
-    "type": "object"
-  },
-  {
-    "additionalProperties": false,
-    "properties": {
-      "fileId": {
-        "description": "ID of an audio file previously uploaded via the Files API (purpose=audio).",
-        "type": "string"
-      },
-      "type": {
-        "const": "file",
-        "type": "string"
-      }
-    },
-    "required": [
-      "type",
-      "fileId"
-    ],
-    "type": "object"
-  }
-]New value: +[
+  {
+    "properties": {
+      "fileUrl": {
+        "description": "HTTPS URL to an audio file (mp3/wav/flac/ogg/webm/m4a).",
+        "type": "string"
+      },
+      "type": {
+        "const": "file_url",
+        "description": "Transcribe audio from the URL in fileUrl.",
+        "type": "string"
+      }
+    },
+    "required": [
+      "type",
+      "fileUrl"
+    ],
+    "type": "object"
+  },
+  {
+    "properties": {
+      "fileId": {
+        "description": "ID of an audio file previously uploaded via the Files API (purpose=audio).",
+        "type": "string"
+      },
+      "type": {
+        "const": "file",
+        "description": "Transcribe an uploaded audio file identified by fileId.",
+        "type": "string"
+      }
+    },
+    "required": [
+      "type",
+      "fileId"
+    ],
+    "type": "object"
+  }
+]

addedInput schema / properties / audio / description

Added value: +"Audio to transcribe, supplied as a public URL or an uploaded file ID."

addedInput schema / properties / contextBias / description

Added value: +"Words or phrases to favor when decoding the audio."

addedInput schema / properties / diarize / description

Added value: +"Identify speakers in the returned transcription segments. Defaults to false."

removedInput schema / properties / model / enum

Removed value: -[
-  "voxtral-mini-latest",
-  "voxtral-small-latest"
-]

addedInput schema / properties / model / maxLength
```
Added value: +200
```
addedInput schema / properties / model / minLength
```
Added value: +1
```

addedInput schema / properties / temperature / description

Added value: +"Sampling temperature for transcription."

changedOutput schema / $schema

Previous value: -"http://json-schema.org/draft-07/schema#"New value: +"https://json-schema.org/draft/2020-12/schema"

addedOutput schema / properties / language / anyOf

Added value: +[
+  {
+    "type": "string"
+  },
+  {
+    "type": "null"
+  }
+]

removedOutput schema / properties / language / type
```
Removed value: -[
-  "string",
-  "null"
-]
```

First observedv0.4.0Apr 23, 2026

TDQS

A4.4/5.0

Behavior4/5

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

Annotations already declare readOnlyHint, idempotentHint, and destructiveHint=false, so the safety profile is covered. The description adds useful behavioral context beyond that: accepted model IDs, default values for diarize and model, the only-supported granularity, and the shape of the response (text, language, segments, token usage). It does not mention auth or rate limits, but that is a minor gap.

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?

The description is front-loaded with the core action, then organizes models, source options, options, and return values into scannable bullet groups. It is efficient, though the option bullets partially duplicate what the schema already documents with 100% coverage.

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

Completeness5/5

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

With an output schema, full annotation coverage, and 100% schema description coverage, the description supplies everything an agent needs: models, input modes, option semantics, and defaults. No material gap remains for invoking this tool correctly.

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 description coverage is 100%, so the baseline is 3. The description nonetheless adds meaning beyond the schema by explaining intent ('language ... boosts accuracy when known', 'contextBias: phrases that should bias the decoder', 'diarize: return per-speaker segments'), which helps the agent use the parameters correctly.

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?

The first sentence gives a specific verb and resource ('Transcribe an audio file to text'), which is unambiguous and clearly distinct from the sibling tools (chat, FIM, vision, OCR, document processing). An agent can immediately tell this is the audio speech-to-text tool without opening any schema.

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?

The description clearly lays out the two mutually exclusive audio source modes (public URL vs. uploaded file ID) and when each applies, which is exactly the routing decision an agent must make. It does not name alternatives or exclusions for when *not* to transcribe, but the context is otherwise clear.

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

Deploy Server