BenedictKing/ccx

Health v2 · maintenance only

Adoption not part of the score

3992 stars · 324 forks observed · 2026-10-08

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-29, confidence not recorded

CCX is a self-hosted AI API proxy and protocol translation gateway written in Go that unifies access to Claude, OpenAI Chat/Images/Embeddings, Codex Responses, and Gemini APIs behind a single endpoint. It ships with an embedded web admin console, multi-key channel management, smart scheduling with health checks and failover, and single-port binary or Docker deployment.

Use cases

• proxy claude api requests through a unified gateway
• translate openai chat completions format to gemini api
• manage multiple api keys with rotation and failover
• route llm requests to channels by model context window
• self-host an ai api gateway with a web admin ui
• add fallback and circuit recovery for llm providers
• expose one endpoint for claude codex and gemini clients

When to choose

• you need one unified endpoint across multiple LLM providers and protocols
• you want self-hosted multi-key management, failover, and health checks for AI APIs
• you prefer a single-binary or Docker deployment with a built-in admin console

When to avoid

• you only use a single LLM provider and need no protocol translation
• you need a fully managed cloud gateway rather than self-hosting
• you require provider features CCX does not proxy, such as fine-tuning or batch jobs

Facets

service · maturity active

api-gateway proxy http-server middleware llm-inference monitoring auth routing large-language-models apis self-hosted developer-tools artificial-intelligence windows self-hosted go api-proxy protocol-translation claude-api openai-api gemini-api multi-key-management failover channel-management web-admin-console single-binary-deployment linux macos docker web-server

Member repositories

For agents

markdown · JSON · MCP: product_card(name="BenedictKing/ccx")