CSM-1B (Conversational Speech Model) is a 1-billion parameter open model capable of generating realistic speech audio conditioned on text and reference audio from specific speakers. It supports multi-speaker conversations and maintains voice consistency across turns. The model is distributed on Hugging Face and is intended for developers building voice interfaces, dubbing tools, and conversational audio applications.
Csm 1b is a Voice, TTS & speech product. It focuses on generating natural conversational speech with consistent voices from text prompts and speaker references. It is built as an open-source project for developers. Csm 1b is open source under the Open Source license. Csm 1b is available on the web and API.
Sesame builds and maintains Csm 1b, and the product first shipped in 2025. Key capabilities include Speech Generation, Voice Cloning, and Conversational Audio. It exposes integrations via a public API.
Latest indexed changes and source events
sesame/csm-1b verified by the PulseGate indexer
Other apps tracked under the same category.
