axonhub

AxonHub is an open-source AI gateway written in Go that sits between application code and model providers, routing requests to over 100 LLMs including Claude, GPT, Gemini, and DeepSeek without requiring SDK or code changes. Developers can point an existing OpenAI SDK or Anthropic SDK at the AxonHub endpoint and immediately access any supported model by adjusting configuration alone. The gateway includes automatic failover with sub-100ms rerouting, weighted load balancing across multiple provider channels, per-request cost tracking that breaks down input, output, and cache token spend, and full request tracing with thread-aware timelines for debugging. Enterprise role-based access control lets teams enforce usage quotas and data isolation across users or projects. Claude Code and other Anthropic-ecosystem tools connect through the standard Anthropic API compatibility layer. The project is particularly useful for teams managing multi-provider LLM infrastructure who want centralized observability and budget controls without maintaining separate integration code for each vendor.

What people ask about axonhub