10 open-source GitHub repos for voice AI: from local ASR to zero-shot voice cloning
Shruti_0810 · x · 2026-11-10
A curated list of 10 open-source GitHub projects for giving AI the ability to listen and speak:
- Whisper.cpp: efficient C/C++ implementation of Whisper for local speech recognition
- Kokoro: lightweight TTS model producing natural-sounding speech (10k stars on GitHub)
- F5-TTS: expressive text-to-speech generation
- Fish Speech: multilingual speech generation and voice cloning
- CosyVoice: multilingual speech with zero-shot voice cloning
The list also includes Oxylabs Web API MCP for connecting agents to web search, covering speech recognition, natural TTS, cloning, and audio understanding for developers experimenting with voice AI.
More from coding & agent
- TensorTonic Launches JAX for ML Systems Study Plan With 35 Hands-On Problems — burny_tech · 2026-11-10
- Anthropic's rules of thumb: start on Opus 5.5, escalate to Fable 5.2 for the hardest work — JeremyNguyenPhD · 2026-11-10
- Tracing a Google ADK multi-agent app with Opik: Flash-Lite 5x cheaper, 6x faster — dl_weekly · 2026-11-10
- Context graphs fix RAG's blind spot: attaching time, audience and provenance to retrieved evidence — JeremyCMorgan · 2026-11-10
- Dev builds pickleball game with Fable 5.2 and Unreal Engine 5.9's AI-first workflow — chongdashu · 2026-11-10
- Google building Agent Swarm mode for Antigravity with shared message board, /swarm command leak claims — testingcatalog · 2026-11-10