AI · AI model
Kyutai TTS (tts-1.6b-en_fr)
Open-weights streaming text-to-speech model with ultra-low latency for English and French
Kyutai TTS is an open-weights, 1.8B-parameter streaming speech model designed for real-time voice applications. It begins generating audio after just a few input words and runs with minimal latency, supporting high-throughput batching on single GPUs. The weights are released under a permissive CC-BY-4.0 license, making it suitable for production conversational agents.
What it's for
- Synthesize real-time voice responses for conversational AI agents
- Stream live text-to-speech outputs in French and English
- Deploy multi-user low-latency voice services on self-hosted GPU hardware
- Apply custom pre-computed voice embeddings for varied speaker styles
Features
- Hierarchical Transformer architecture (1B backbone, 600M depth transformer) using Mimi audio tokens
- Low-latency streaming generation that starts producing audio before reading the entire prompt
- Trained on 2.5 million hours of audio with CFG distillation for faster inference
- Supports concurrent batching up to 75x realtime generation throughput
- Voice conditioning via pre-computed cross-attention embeddings
- Bilingual support for English and French speech
Pricing
Open source. Free, open-source under CC-BY-4.0
Details
- Type
- AI model
- Category
- AI
- Topic
- Text to Speech
- Open source
- Yes
- Platforms
- Linux, Self-hosted, Code library
- For
- Developers and AI researchers building real-time conversational voice products
- Alternative to
- ElevenLabs, OpenAI TTS, Cartesia, Kokoro TTS
- Tags
- tts, streaming audio, voice ai, moshi, speech synthesis, open weights
Why it was saved
- @reach_vb: Open-license, low-latency streaming TTS for building real-time voice products.
Compare
Related
AI voice generator, voice cloning, and conversational agent platform
Ultra-low-latency streaming text-to-speech and instant voice cloning model.
Open-source 1.6B text-to-speech model generating multi-speaker dialogue in one pass.
Fast, low-cost text-to-speech API and studio for creators and developers
Converts spoken thoughts into structured, polished writing in customizable styles.
Official Python library for Cartesia's real-time text-to-speech API
Updated