Moshiko (part of the Moshi project by Kyutai) provides an open-source PyTorch implementation of a speech/audio codec and model. It enables encoding and decoding of 24kHz mono audio using the Mimi codec, with support for running an interactive local web server. The model is hosted on Hugging Face and is designed for researchers and developers building local audio AI applications.
Moshiko is an Other AI product. It focuses on running high-quality open speech and audio models locally without proprietary APIs. It is built as an open-source project for AI researchers and developers. Moshiko is open source under the Apache-2.0 license. Moshiko is available on the web and the command line, and it can be self-hosted.
Behind Moshiko is Kyutai, based in France, and the product first shipped in 2024. Development happens publicly on GitHub with 10.6k stars and 2 commits in the last 90 days. Key capabilities include Audio Encoding, Audio Decoding, and Interactive Web Server.
Latest indexed changes and source events
kyutai/moshiko-pytorch-bf16 verified by the PulseGate indexer
Other apps tracked under the same category.
