Ir para o conteúdo
AI.info

AI Tools

Moshi

Moshi: preços, capacidades, atualizações e alternativas, lidos nas páginas do próprio desenvolvedor.

O site de Moshi

Em inglês

Key facts

Price
Free Read on kyutai.org
Free plan
Not stated on the maker’s pages
Platforms
Web · macOS · Linux · CLI
API
Yes
Open source
Yes · MIT Read on github.com
Latest version
mlx-0.3.0
Last update
8 Jan 2026 Read on github.com
Prices checked
10 Nov 2026 Read on kyutai.org

Moshi listens and speaks directly through audio streams instead of using separate speech recognition, text generation, and text-to-speech stages. It models the user’s and assistant’s audio in parallel and produces an inner-monologue text stream to guide responses.

Researchers and developers can run it locally through PyTorch, MLX, or Rust backends, use its web interface, or adapt the code for experiments. It speaks English and requires substantial hardware for some configurations; Windows is not officially supported.

Category
Audio & Music
Read from the maker’s pages
2026-09-19

Company

Kyutai

Features

  • Full-duplex spoken dialogue with simultaneous listening and speaking
  • Direct speech-to-speech generation with parallel audio streams
  • Inner-monologue text stream to improve response generation
  • Streaming Mimi neural audio codec for 24 kHz audio
  • PyTorch, MLX, and Rust/Candle inference backends
  • Local web interface and command-line client
  • Three released model variants, including Moshika and Moshiko
  • Code under MIT and Apache-2.0 licenses; model weights under CC-BY 4.0

Use cases

  • Build low-latency voice conversation prototypes
  • Research speech-native language models
  • Run voice dialogue locally on Mac or Linux hardware
  • Fine-tune voice dialogue behavior for experiments
  • Create custom web or command-line speech interfaces

Pricing

Free

Prices checked
2026-11-10

Free / Demo

free

  • Real-time full-duplex voice conversation (sub-200ms latency)
  • Experimental conversational AI prototype (non-commercial use)
  • Limited to 5-minute conversations
  • Open-source models (Moshi, Mimi codec) available for self-hosting
  • Access via online demo at moshi.chat / kyutai.org

Latest updates

Where it works

Web · macOS · Linux · CLI

More from Kyutai

  • Unmute

    Unmute adds real-time speech input and output to text-based LLMs using Kyutai speech models.

Questions about Moshi

How much does Moshi cost?

The plans on the maker’s pricing page: Free / Demo (free). Prices checked on 10 Nov 2026.

Does Moshi have an API?

Yes. The maker’s pages state that Moshi has an API. Read on the maker’s pages on 19 Sept 2026.

Which platforms does Moshi run on?

Moshi runs on Web, macOS, Linux and CLI, as its maker’s pages state. Read on the maker’s pages on 19 Sept 2026.

Is Moshi open source?

Yes. The code of Moshi is published at github.com/kyutai-labs/pocket-tts under the MIT licence. Read on the maker’s pages on 19 Sept 2026.

Is Moshi still available?

Yes, as of 19 Sept 2026: its maker’s pages were read that day and carried no notice that Moshi has ended.

What changed recently in Moshi?

The latest entry in its maker’s changelog: 8 Jan 2026, “moshi-v0.2.12”. Before it: 4 Aug 2025, “mlx-0.3.0: Add batching support to the MLX implementation (#347)”. Changelog checked on 15 Nov 2026.

Sources

Not confirmed

Prices
No official Moshi pricing page was found after searching for Moshi pricing.
No official pricing page was available to read.
No official pricing tier was published on a page found during this search.
No official pricing page was found, so pricing tiers are omitted.
Free plan
Pricing and subscription terms were not stated on the official pages reviewed.
Official website