🔥The 9th Round of Easy Loan, Earn $40 Reward is in progress❗️ ⏰ Promotion Period: January 15th - Feburary 15th, 2025 👉 Register now and check more details at gate.io/campaigns/358

Want to contribute to the future of #AI? Our lab is offering #research internships in 2025, providing a unique opportunity to work on groundbreaking projects. Don't miss out 👀! #internship kyutai.homerun.co/research-inter…

Had a great time at the 'AI on the Edge' hackathon this afternoon, huge kudos to the organizers spotlight ! I was super impressed to see how much teams built in just two days, truly a testament to the power of open-source for AI innovation! 🚀 #AI #OpenSource

Excited to talk at #RustNation25 next Feb! Join us to discover how we leveraged #rust 🦀 to improve a mostly Python based AI research workflow and create Moshi 🟢the first open-source speech-to-speech model.

I’m honored that SoundStream (2021) retrospectively receives the best paper award from IEEE Signal Processing Society. We introduced neural audio codecs, with telecom in mind, and realized later that it allowed for creating audio language models that are now a standard framework for audio synthesis.

Super proud of our first publicly released text model, helium-1 preview, a 2B model trained on 6 languages. It should be a great fit for on-device applications. Already available in candle/transformers, can't wait to see what the community builds with it! #OpenSource #AI #FTW !

Helium 2B running locally on an iPhone 16 Pro at 28 tok/s, faster than you can read, all thanks to mlx-swift with q4 quantization!

Very impressive to hear this Japanese 🇯🇵 version of moshi 🟢. I don't speak the language so I cannot understand what it's trying to tell me but at least it sounds great 😅

Super happy that we're releasing hibiki today, our first speech 🇫🇷 to speech 🇬🇧 translation model. It works in real-time while preserving the voice of the speaker, and best of all it can run on your phone. Code on GitHub (pytorch, mlx, swift, rust), and weights on HF 🚀

Neil showing Hibiki 🟢 running on an iphone. All local so whatever you say will stay on the iphone and latency is great! Get the code to build the ios app in our moshi-swift repo (it's brittle at the moment but we'll polish it 😅 )

Afraid of missing out on French pop culture references because you don't speak the language? Fear no more and try our Hibiki speech-to-speech translation model— no more FOMO! 🇫🇷✨ #Translation #AI

Here one kyutai speech-to-speech Hibiki example with Xavier Niel 🤗

Thanks to Xavier Niel for stopping by at the #AIActionSummit to try Hibiki. No need to struggle with English anymore 😅

Want to speak English in the club even in very loud conditions 🎤🎵🎶, bring Hibiki with you in your iphone📱 Even works if you don't have some network access!

What are we waiting for? 🤔

Want to train your own Moshi🤖? You'll find the code to do that on github!

We've just released Helium 1, a 2B model that is best of its class on multi-lingual benchmarks and supports the 24 EU languages!

🚀 Say hello to unmute.sh — a modular voice AI system built on our in-house low latency text-to-speech and speech-to-text engines. It works in English🇬🇧and French 🇫🇷 and you can customize the voice and personality 🎙️Try it live and tell us what you think!

Kyutai Speech-To-Text is now open-source! It’s streaming, supports batched inference, and runs blazingly fast: perfect for interactive applications. Check out the details here: kyutai.org/next/stt