---
title: Voxtral Realtime 4B
emoji: 🎙️
colorFrom: yellow
colorTo: red
sdk: static
pinned: false
license: apache-2.0
short_description: Speech-to-Text in the browser with transformers.js + WebGPU
language:
  - en
  - fr
  - es
  - de
  - ru
  - zh
  - ja
  - it
  - pt
  - nl
  - ar
  - hi
  - ko
---

# Voxtral Realtime 4B — Live Speech-to-Text

Real-time speech transcription running entirely in your browser using [Voxtral-Mini-4B-Realtime](https://huggingface.co/mistralai/Voxtral-Mini-4B-Realtime-2602) via [transformers.js](https://github.com/huggingface/transformers.js) + WebGPU.

- Click the mic to start listening
- VAD automatically detects speech segments
- Words appear as the model generates them
- All processing happens locally — no server needed