A dependency-free AI chat assistant built in vanilla HTML, CSS, and JavaScript — no framework, no backend. It integrates the Groq API for LLM inference (Llama 3.3 70B) and the Web Speech API for voice input and output, achieving sub-second responses on Groq’s LPU-accelerated hardware. Deployed on GitHub Pages, under 50KB, zero server costs.
Category:
Voice AI / LLM
Client:
Personal Project
Stack:
Vanilla JS · Groq · Web Speech API
Location:
Chennai, India
Built By:
R JAN STEVE DANIEL
