ourongxing/chatgpt-vercel

Built on Solid-Start framework with client-side rendering, it features multi-conversation support with independent per-session configuration, prompt presets with fuzzy search, real-time token usage tracking, and PWA capabilities. The architecture supports multiple OpenAI API keys for load balancing, customizable model parameters (temperature, context window), and edge function deployment on Vercel with region-specific routing to avoid API restrictions. It integrates with OpenAI's API, supports local session persistence with optional cloud export/import, and can be deployed across multiple platforms (Vercel, Netlify, CloudFlare Workers, Railway).

3,231 stars. No commits in the last 6 months.