How I Built a Free AI Chatbot With 12+ Models, Voice, Coding, and Image Generation — All for $0

Description and use

A free AI assistant that combines chat, web search, code execution, image generation, and voice using 12+ free models and Convex for real‑time interactivity.

Problem

Develop a cost‑effective AI assistant that offers advanced features (chat, web search, code execution, image generation, voice, file analysis, multi‑agent collaboration) without subscription fees.

Outcome

Free AI provides a fully functional, subscription‑free AI assistant with advanced capabilities, achieving 99.9% uptime through multi‑provider fallback and real‑time interactivity via Convex.

Workflow

• 1. Set up React + TypeScript frontend with Vite, Tailwind, and Framer Motion.
• 2. Configure Convex backend for real‑time database, serverless functions, and email OTP auth.
• 3. Integrate 12+ free AI models via NVIDIA NIM, Groq, Google Gemini, and OpenRouter.
• 4. Implement model routing: try NVIDIA models, fallback to Groq, then Gemini Flash, then OpenRouter, each with 12‑second timeout.
• 5. Add code execution using Pyodide for Python and sandboxed JavaScript.
• 6. Enable voice with Edge TTS and Web Speech API.
• 7. Add web search using Exa API with source citations.
• 8. Build multi‑agent bot system: planner agent creates task graph, assigns roles (PM, frontend dev, backend engineer, QA, security reviewer), shares workspace, and streams activity to user.
• 9. Deploy and expose free access without signup.
• 10. Iterate based on community feedback.

Evidence

Free AI — a completely free AI assistant that does what paid tools charge $20/month for.

Source

The system automatically routes requests to the best available model: NVIDIA, Groq, Gemini Flash, OpenRouter.

Source