Skip to content

feat: add transcription support for Mistral (Voxtral) #1420

@renanvieira

Description

@renanvieira

Feature Request

Add transcription support for Mistral Voxtral models.

Motivation

Mistral offers a transcription API for their Voxtral models (voxtral-mini-latest, voxtral-small-latest) but Rig doesn't support it yet. Other providers (OpenAI and Gemini) already have transcription implementations.

Proposal

Implement the TranscriptionModel trait for the existing Mistral provider, following the same pattern as existing implementations.

Request to POST /v1/audio/transcriptions with the audio file and some parameters like language, diarization and others. The response types includes text and extra ,optional, information like diarization segments and token usage stats.

Full API documentation: https://docs.mistral.ai/api/endpoint/audio/transcriptions

Alternatives

None — this is a straightforward trait implementation with other examples already in the codebase.

Metadata

Metadata

Assignees

No one assigned

    Labels

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions