FlashLabs-Chroma
Open-source real-time spoken dialogue with personalized voice cloning.
Moshiは、kyutai-labsが公開しているオープンソースの音声プロジェクトです(GitHub: kyutai-labs/moshi)。GitHubスターは11,363、フォークは1,051で、Apache-2.0ライセンスの下で公開されています。GitHubでの最終更新は2026年9月9日です。
Moshi — Kyutai full-duplex spoken dialogue model: talk and listen at the same time, ~200ms latency, Apache-2.0
Install the PyTorch package from PyPI, or the MLX package on Apple Silicon.
Run the Moshi server, optionally with a Gradio tunnel, pointed at the Moshika weights.
Run the local MLX client with 4-bit or 8-bit quantisation, or connect the Python client to your server.
Speaking and listening happen simultaneously — interruptions and overlap work the way real conversation does.
Mimi streams 24 kHz audio down to a 12.5 Hz representation at 1.1 kbps, with only 80ms of frame latency.
Around 200ms overall on an L4 GPU — fast enough for natural spoken interaction.
A PyTorch server path and an MLX local path with quantised weights for Macs.
The README install route from PyPI.
pip install -U moshiThe README MLX route with 4-bit quantised weights.
python -m moshi_mlx.local -q 4The README server route, with the Moshika PyTorch weights.
python -m moshi.server --hf-repo kyutai/moshika-pytorch-bf16Kyutai, as a speech-text foundation model and full-duplex dialogue framework.
Yes — Apache-2.0 licensed.
The model listens while it speaks, so you can interrupt it and it can overlap — unlike turn-based voice assistants that wait for silence.
Theoretical latency 160ms, practical overall latency as low as 200ms on an L4 GPU, via the Mimi streaming codec.
pip install moshi (or moshi_mlx on Apple Silicon), then run the server or the local client.
Open-source real-time spoken dialogue with personalized voice cloning.
Clone anyone's voice and personality for real-time conversation.
A generative speech model built for daily dialogue — natural, conversational text-to-speech
MOSS-TTS-Nano: a 0.1B open multilingual speech model that clones voices and streams in real time on CPU
A TTS model capable of generating ultra-realistic dialogue in one pass.
MOSS-TTS family: open text-to-speech models for voice cloning, dialogue, voice design, and real-time agents
LibreAIカタログに掲載された音声プロジェクトの中で最もスターの多いもの — Moshiが合わなかった場合に検討したい、現実的な候補リストです。
利用 AI 大模型和自动化工作流,根据主题或关键词一键生成高清短视频。Generate HD short videos from a topic or keyword with an automated AI workflow.
Robust Speech Recognition via Large-Scale Weak Supervision
1 min voice data can also be used to train a good TTS model! (few shot voice cloning)
Clone a voice from 5 seconds of audio and generate arbitrary speech in real time
Port of OpenAI's Whisper model in C/C++
カタログの中でMoshiとスター数が近い音声プロジェクト — 前後の棚もそのまま見て回れます。
A PyTorch-based Speech Toolkit
Self-hosted backend that gives ESP32 'Xiaozhi' devices streaming voice chat, memory and tool control.
Use Microsoft Edge's online text-to-speech service from Python WITHOUT needing Microsoft Edge or Windows or an API key
Neural building blocks for speaker diarization: speech activity detection, speaker change detection, overlapped speech detection, speaker embedding
Easy-to-use Speech Toolkit including Self-Supervised Learning model, SOTA/Streaming ASR with punctuation, Streaming TTS with text…
Silero VAD: pre-trained enterprise-grade Voice Activity Detector