VoidLLM is a self-hosted LLM proxy and AI gateway that routes requests across OpenAI, Anthropic, Azure, Ollama, and vLLM — with load balancing, failover, API key management, usage tracking, and rate limiting. What makes it different: zero-knowledge architecture. VoidLLM never stores or logs any prompt or response content. Not a toggle — by design. GDPR-compliant from day one. Built in Go for sub-2ms overhead. Single binary, full web UI included. Free tier with unlimited users.
Hey PH! I built VoidLLM because every LLM proxy I tried either logged my prompts for "observability" or added 10ms+ overhead with Python.
VoidLLM is a Go-based LLM gateway that sits between your apps and any LLM provider. It handles auth, rate limiting, cost tracking, and load balancing across multiple deployments — but it never touches your prompt data. Zero-knowledge by architecture, not by configuration.
The free tier includes almost everything except SSO and audit logs. Happy to answer any questions!
Report
Reviews
No reviews yetBe the first to leave a review for VoidLLM