speech-to-speech

What it does

A low-latency VAD → STT → LLM → TTS pipeline behind an OpenAI Realtime-compatible WebSocket API, where every stage is replaceable — point the LLM slot at a hosted provider, HF Inference Providers, or your own vLLM or llama.cpp server for a fully local stack. Latency and quality depend entirely on the parts you pick, so run the defaults first. Published by Hugging Face and used in production. Apache-2.0.

Summary based on the project's own description. The Japanese page has a fuller beginner guide: setup difficulty, what to prepare, and a step-by-step start.

Screenshots & visuals

Before you start

• Setup difficultyEasy to start
• Command lineRequired
• API keyNot required
• Sends data onlineNo

Judged automatically from the project's own metadata. Always confirm on the official page.

speech-to-speech FAQ

Is speech-to-speech free?

Yes. It runs on your own machine, so there is no usage-based cost either.

Do I need to install anything or write code to use speech-to-speech?

It needs the command line. If you are not comfortable typing commands, expect to get stuck during setup.

Does speech-to-speech work on a phone?

No. It needs a computer.

Can I use speech-to-speech commercially?

The licence is Apache-2.0, which permits commercial use. Conditions such as keeping the copyright notice still apply, so read the licence text before shipping it inside something.

How it compares

The conditions that decide whether you can actually use it, side by side.

"Sends data out" means your input is uploaded to someone's server. Pick "No" when the material cannot leave your organisation.

Other ai agents tools