Open-source voice AI platform where every component is swappable — any LLM (OpenAI, Anthropic, Groq), any TTS with a free local option, any STT. No lock-in, MIT licensed. Includes a live web canvas where the AI generates full HTML pages during conversation — dashboards, reports, interactive tools — in a desktop-style UI. Plus AI image gen, voice cloning, agent profiles, and AI music. One-click local instal via https://pinokio.co/ Deploy on any VPS: npx openvoiceui setup

Framer — Launch websites with enterprise needs at startup speeds.

Launch websites with enterprise needs at startup speeds.

I built OpenVoiceUI because every voice AI tool I tried locked you into one provider. Change your LLM? Rewrite your app. TTS service changes pricing? Start over. The idea was simple: fully separate the interface from the intelligence. Every component — LLM, TTS, STT — is swappable without touching anything else. It ships with a free local TTS engine (Supertonic) so you can run it for zero API cost. The part I'm most excited about is the live web canvas. During a voice conversation, the AI can generate and render full HTML pages in real-time — dashboards, reports, interactive tools — displayed in a desktop-style UI with windows and folders. It's MIT licensed, Docker-based, and installs with one command: npx openvoiceui setup Would love feedback on the canvas system and the pluggable architecture. Happy to answer any questions!

Framer — Launch websites with enterprise needs at startup speeds.

Launch websites with enterprise needs at startup speeds.

OpenVoiceUI

Voice AI platform. Any LLM. Any TTS. No lock-in.

Voice AI platform. Any LLM. Any TTS. No lock-in.