This is an early release preview. You may encounter bugs.
AxonHub logo
Unclaimed

Tool coding interface

AxonHub

AI gateway that translates any SDK's requests to any provider, with tracing, RBAC, load balancing, and cost tracking

A+ 94/100 GitHub score ? This grade is derived from GitHub signals, not user votes. Open for the full breakdown.
No votes yet

01 / About

What AxonHub is.

AxonHub is a self-hosted AI gateway that sits between your application and model providers. It converts requests between SDK formats, so code written against the OpenAI SDK can call Claude or Gemini and code written against the Anthropic SDK can call GPT models: you point the SDK's base URL at AxonHub and change the provider in configuration rather than in code.

Providers are added as channels, each holding an API key and the set of models it serves. Client applications authenticate with an AxonHub API key instead of a provider key, and the gateway routes to the healthiest channel, failing over in under 100 milliseconds when one degrades.

Beyond routing, it records a thread-level timeline for every request, tracks cost per request across input, output, and cached tokens, and enforces role-based access control with usage quotas and data isolation. The server is written in Go and ships as a Docker image with a web console for channels, models, model prices, traces, and request monitoring.

API type Status
Text generation (OpenAI, Anthropic, and Gemini endpoints) Available
Image generation Available
Rerank Available
Embedding Available
Realtime conversation Planned

Features

  • SDK-to-model translation: any supported SDK format reaches any supported provider without code changes
  • Provider coverage: OpenAI, Anthropic, Zhipu, Moonshot, DeepSeek, Doubao, Gemini, Jina AI, OpenRouter, ZAI, and NanoGPT, with AWS Bedrock and Google Cloud in testing
  • Channel management: each channel holds a provider key and its enabled models, with connection tests before enabling
  • Smart load balancing: automatic failover under 100 ms routes to the healthiest channel
  • Request tracing: a full per-request timeline at thread-level granularity
  • Cost tracking: per-request cost breakdowns covering input, output, and cached tokens
  • Access control: role-based permissions, usage quotas, and data isolation between users
  • API surface: OpenAI, Anthropic, and Gemini compatible endpoints plus image generation, rerank, and embedding APIs
  • Deployment: Docker-ready Go binary with an initialisation wizard for the first administrator account

02 / Discussion CREDIBILITY-GATED

Discussion

Reading is open to everyone. Posting and voting need a verified identity or a GitHub grade of B or higher.

  • No discussions yet.

04 / Build

Build with AxonHub.

Browse the catalogue for frameworks, tools, and harnesses, each scored on real GitHub credibility.

Get AxonHub →

Browse the catalogue