billiax/voxglide
Embeddable voice AI SDK for web pages — speak to fill forms, click buttons, navigate, and ask questions. Works with any LLM.
Combines browser Speech API with WebSocket-based LLM integration through a server proxy that holds API keys, auto-discovers DOM elements (forms, buttons, navigation) to build page context, and exposes custom tools via `window.nbt_functions`. Built with zero dependencies and Shadow DOM isolation, it supports Gemini, OpenAI, Anthropic, and Ollama with multi-step conversation workflows and accessibility features like ARIA live regions.
Available on npm.
Stars
5
Forks
—
Language
TypeScript
License
MIT
Category
Last pushed
Mar 19, 2026
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/voice-ai/billiax/voxglide"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Higher-rated alternatives
Thiagohgl/ai-pronunciation-trainer
This tool uses AI to evaluate your pronunciation.
alexiokay/AriLink
Modern ARI-STASI server, built on Asterisk ARI with real-time speech-to-text transcription,...
rohit-lakhanpal/ai-hackathon-starter-kit
This project has been created to make AI accessible and easy for everyone. Whether you are a...
mikopbx/ModuleSmartIVR
Модуль умной маршрутизации для 1C:Предприятия
matt-goldman/AI-Panelist
An AI Panelist participating in Beer Driven Devs Live 2026