voxtral

Voxtral ASR & TTS running natively and in the browser. A Rust implementation of Mistral's Voxtral mini realtime ASR / TTS using the Burn ML framework

Turn PDFs and EPUBs into audiobooks; subtitles or videos into dubbed videos (including translation), and more. For free. Pandrator uses local models, including voice-cloning (instant, RVC-enhanced, XTTS fine-tuning) and LLM processing. It aspires to be a user-friendly app with a GUI, an installer and all-in-one packages.

C++ ggml runtime hub for multilingual ASR and TTS models: Cohere Transcribe, Parakeet TDT, Voxtral, Canary 1B v2, etc, plus universal forced alignment, and more

Super STT let's you speak, and your words are typed straight into whatever app is focused

Offline Speech-to-Text (STT) service using Mistral's Voxtral model with Wyoming protocol compatibility for Home Assistant Assist integration.

Voxtral is a state-of-the-art model developed to handle both speech transcription and audio understanding with remarkable accuracy and efficiency. This demo interface lets you run the Voxtral model on powerful GPUs to evaluate its performance and see how it can be used for transcription and deeper analysis.

Effortless Push-to-Talk Transcription, Anywhere.

speech to text gui for different (e.g. Whisper, Voxtral) models and backends, including whisper.cpp, crispasar, mlx-whisper, faster-whisper, ctranslate2; applies pyannote for diarization

A Web UI for easy subtitle using various models including voxtral

Professional local-first AI production pipeline for long-form narration. Clone voices and generate studio-grade audiobooks (M4B/MP3) using Coqui XTTS-v2 and support for Voxtral (cloud)

Voxtral Codec : Combining Semantic VQ and Acoustic FSQ for Ultra-Low Bitrate Speech Generation (Voxtral TTS Backbone)

Experimentation with Voxtral-Mini-4B-Realtime-2602 and DeepL API for live translation

Talk. Ink. Push-to-talk dictation for macOS, 100% on-device. Pick your model: Qwen3-ASR, NVIDIA Nemotron or Voxtral, all via Apple MLX.

github mirror for radioshaq - ham radio full time quarterback and part-time lobster

Real-time face-to-face translation app (React Native)

Live VRChat voice translator with OSC chatbox output and your choice of AI provider.

Real-time phone scam detection powered by Mistral's Voxtral Mini - analyzes live audio and transcripts to identify fraud patterns

Local implementation for voxtral

Enterprise-grade speech-to-text toolkit with pluggable backends (Whisper, Voxtral). Features speaker diarization, 80%+ test coverage, CI/CD quality gates, and fully offline operation.

Open-source, local-first, system-wide voice dictation for Windows