Izwi 0.1.0 ships fully local voice AI with TTS, cloning and diarization
zinyando · reddit · 2026-10-26
Izwi 0.1.0, the first stable release of an Apache-2.0 open-source voice AI app for Mac, Linux and Windows, is out. It bundles local ASR → LLM → TTS real-time voice chat, voice cloning from short samples, text-described voice design, a Studio for long-form audio, transcription with speaker diarization and forced alignment, and an OpenAI-compatible /v1 API so existing code can point at localhost.
Supported models include TTS (Qwen3-TTS, Kokoro-82M, Fish Audio S2 Pro, VibeVoice, Voxtral TTS, LFM2.5 Audio), ASR (Parakeet TDT v3, Whisper large-v3-turbo, Qwen3-ASR, Granite Speech 4.1, Nemotron, Voxtral Mini), diarization (Sortformer, Nemotron 3) and chat (Qwen3.5/3.6/3.8, LFM2.5). Inference runs natively in Rust on Candle with custom Metal and CUDA backends — no Python environment needed.
Related event: Izwi 0.1.0 launches as all-local voice AI app(2 posts)→
More from Infra
- Kimi K3's ~3T parameters spread across 59 Cerebras WSE-3T wafers, weights stay resident — petrusenko_max · 2026-10-26
- OpenAI's agent built the same site 8 times — every deployment landed on Cloudflare — FinanceYF5 · 2026-10-26
- DIY LLM-designed 3D-printed chassis housing four AMD Instinct v620 accelerators — PraxisOG · 2026-10-26
- Building AI infra may be easier than selling AI outcomes with clear ROI — seankinneyRCR · 2026-10-26
- Frontier security models will need new compute classes, quantum likely first — BenBajarin · 2026-10-26
- DesignArena data: Gemini 3.8 Flash quality degrades sharply weeks after launch — legit_api · 2026-10-26