Skip to main content

NextChat (ChatGPT Next Web) Custom OpenAI API Base URL Setup Guide (2026)

· 4 min read
AI Supermarket
AI Supermarket
AI API Gateway Docs

NextChat (formerly known as ChatGPT Next Web) is one of the most popular open-source, cross-platform ChatGPT web clients available for self-hosting via Docker or Vercel. However, developers and self-hosters frequently encounter rate limit errors (429 Too Many Requests), high token costs, or lack of multi-model support when relying on default API endpoints.

By overriding NextChat's default API endpoint with a unified OpenAI-compatible API gateway like AI Supermarket, you can unlock unthrottled concurrency, access Claude 3.5 Sonnet, DeepSeek V3, GPT-4o, and Gemini 2.5 under a single API key, and eliminate monthly subscription overheads.

Why Custom Base URL with AI Supermarket Gateway?​

1. Multi-Model Access in NextChat

NextChat is natively designed around OpenAI REST API specifications. By setting your custom base URL to https://aisupermarket.work/v1, your NextChat instance can seamlessly switch between claude-3-5-sonnet-20241022, deepseek-chat, gpt-4o, and gemini-2.5-pro without needing separate API keys for each provider.

2. Bypass RPM & TPM Rate Limits

Official API endpoints enforce strict Tier 1 rate limits. AI Supermarket routes your requests through high-availability enterprise proxy pools, guaranteeing high throughput even for team-wide NextChat deployments.

Docker Environment Variables (Recommended)​

docker run -d \
  -p 3000:3000 \
  -e OPENAI_API_KEY="sk-your-ai-supermarket-token" \
  -e BASE_URL="https://aisupermarket.work/v1" \
  --name nextchat \
  yidadaa/chatgpt-next-web