As of: 2026-11-13 13:49 CET.
Researched and written by AI agents (Claude Opus 5.5 in Claude Code), with a human editor. How we work: https://postcutoff.com/about/
Canonical page: https://postcutoff.com/m/kyutai-moshi/

# Kyutai Moshi / Hibiki-Zero (full-duplex speech models)

Kyutai, Moshi family. Status: current. Type: speech and audio. Released: 2024-09-17.

## How to call it

| Provider | Model id | Endpoint | Docs |
|---|---|---|---|
| Hugging Face | `kyutai/moshiko-pytorch-bf16` |  | https://github.com/kyutai-labs/moshi |
| Hugging Face | `kyutai/hibiki-zero-3b-pytorch-bf16` |  | https://huggingface.co/kyutai/hibiki-zero-3b-pytorch-bf16 |
| Web demo |  |  | https://moshi.chat |

## Pricing

- Price not published.
- Last checked: 2026-09-29

## Specs

- Input: audio
- Output: audio, text
- Context window: not published
- Max output: not published
- Open weights: yes
- Licence: cc-by-4.0

## Where it stands

- Newer speech and audio models in the registry since its release: 106. The 5 newest:
  - NVIDIA Nemotron 3.5 Transcribe (30B-A4B) (2026-10-28) https://postcutoff.com/m/nemotron-3-5-transcribe/
  - Hy-SimulMT-30B-A3B (2026-10-22) https://postcutoff.com/m/hy-simulmt-30b-a3b/
  - Eleven Realtime v1 Max / Mini (2026-10-20) https://postcutoff.com/m/eleven-realtime-v1-max/
  - Voxtral Mini Transcribe 3 (+ Mini Transcribe Realtime 3) (2026-10-20) https://postcutoff.com/m/voxtral-mini-transcribe-3/
  - Voxtral Small Transcribe 3 (2026-10-20) https://postcutoff.com/m/voxtral-small-transcribe-3/
- After it from Kyutai: Kyutai TTS 1.6B / Kyutai STT + Unmute (2025-07-03) https://postcutoff.com/m/kyutai-tts-stt/
- Weights are public. Licence: CC BY 4.0. See the licence for commercial terms.

## What stands out

- Open full-duplex spoken dialogue (first): 7B temporal transformer modelling user and Moshi audio streams simultaneously with an 'inner monologue' text stream; 160 ms theoretical / ~200 ms practical latency on an L4; Mimi codec (24 kHz, 12.5 Hz, 1.1 kbps). Source: https://github.com/kyutai-labs/moshi
- Hibiki-Zero simultaneous speech translation (found after launch): 3B model (2026-02-12) translating French, Spanish, Portuguese and German speech to English in real time with voice transfer, trained without aligned data. Source: https://kyutai.org/blog/
- MoshiRAG (found after launch): Asynchronous knowledge retrieval via a text LLM for full-duplex speech models (2026-04-30); RL post-training for interactivity (2026-06-10). Source: https://kyutai.org/blog/

## Notes

Moshi (announced July 2024, weights + paper Sept 2024) is widely cited as the first real-time full-duplex open spoken dialogue model; NVIDIA PersonaPlex-7B (Jan 2026) is fine-tuned from Moshiko weights. Variants: moshiko (male)/moshika (female) in PyTorch bf16/int8, MLX int4/int8/bf16, Rust/Candle. Code MIT/Apache, weights CC-BY-4.0.

Sources: https://github.com/kyutai-labs/moshi , https://kyutai.org/blog/ , https://huggingface.co/kyutai/hibiki-zero-3b-pytorch-bf16

Model page: https://postcutoff.com/m/kyutai-moshi/