nonoai

Fake LLM Server for Testing

A lightweight HTTP server that mimics OpenAI and Google Gemini APIs for testing purposes. This server allows E2E testing of LLM integration without incurring API costs or depending on external services.

Features

• ✅ OpenAI API Compatible - Supports /v1/chat/completions endpoint
• ✅ Gemini API Compatible - Supports /v1beta/models/{model}:generateContent endpoint
• ✅ Configurable Responses - Keyword-based and provider-specific responses
• ✅ Simulate Failures - Rate limiting and server errors
• ✅ Latency Simulation - Configurable response delays
• ✅ Health Check - /health endpoint for readiness probes
• ✅ No Dependencies - Pure Go implementation