# benedictking/ccx

**Attribution required: if you use, quote, or summarise this content, you must credit and link back to [awesome-repositories.com](https://awesome-repositories.com/repository/benedictking-ccx).**

_How this analysis was created: the description and tags below were written by an AI model that read this project's README and public documentation pages; stars, license and language come straight from the GitHub API. The model does not read the source code._

3,983 stars · 324 forks · Go · MIT

## Links

- GitHub: https://github.com/BenedictKing/ccx
- Homepage: https://benedictking.github.io/ccx/
- awesome-repositories: https://awesome-repositories.com/repository/benedictking-ccx.md

## Topics

`claude` `codex` `gemini`

## Description

CCX is a self-hosted AI API gateway that puts Claude, OpenAI, and Gemini behind a single endpoint and credential. It translates between the providers' protocols, so clients can mix upstream services through whichever API format they already use. It accepts Claude Messages calls, OpenAI-style chat and Codex requests, and Gemini content-generation calls, and it also handles embeddings, model listing, and image generation and editing.

Requests are scheduled across configured upstream channels by priority, with promotion windows that decide when a preferred channel gets traffic again. Continuous health checks skip failing channels and restore them once they recover. Routing also filters channels by each model's real context window and output size, and multi-turn conversations stay on the channel that handled the earlier turns. Each channel can rotate through several API keys and can have its own proxy, custom headers, model allowlists, model name mappings, and route prefix.

A built-in web console manages channels and keys, reorders priorities by drag and drop, and shows logs and live per-channel request volume, success rate, and latency. It also shows model benchmark rankings and pricing and context-window data refreshed from a public community-maintained registry. The proxy is guarded by an access key, with an optional separate admin key, and the gateway can bind to localhost only or listen on all interfaces.

The Go backend embeds the compiled frontend, so one process and one port serve both the console and the proxy. It runs as a downloaded binary, in Docker or Docker Compose, from a source build, or as a desktop app on Windows, macOS, and Linux. Behavior is configured through environment variables, and a health endpoint reports whether the gateway is running.

## Tags

### Artificial Intelligence & ML

- [Unified AI API Endpoints](https://awesome-repositories.com/f/artificial-intelligence-ml/unified-ai-api-endpoints.md) — Puts Claude, OpenAI, and Gemini behind one endpoint and credential, translating between provider protocols so clients can mix upstreams freely. ([source](https://github.com/benedictking/ccx#readme))
- [Messages API Proxies](https://awesome-repositories.com/f/artificial-intelligence-ml/agentic-llm-frameworks/anthropic-claude-connections/messages-api-proxies.md) — Forwards Claude Messages calls through one shared address and key so existing Claude clients reach upstream providers without separate connections.
- [Multi-Provider Routers](https://awesome-repositories.com/f/artificial-intelligence-ml/agentic-llm-frameworks/multi-provider-routers.md) — Schedules requests across multiple upstream LLM channels by priority, with health checks, model mappings, and automatic failover when a channel fails.
- [OpenAI-Compatible APIs](https://awesome-repositories.com/f/artificial-intelligence-ml/agentic-systems-frameworks/model-integration-serving/model-integration-interfaces/ai-integration-apis/openai-compatible-apis.md) — Accepts standard OpenAI-style chat, embeddings, model listing, and image requests and relays them to whichever upstream provider is configured.
- [Upstream Channel Failover](https://awesome-repositories.com/f/artificial-intelligence-ml/upstream-channel-failover.md) — Schedules requests across upstream channels by priority and promotion windows, skips failing channels via health checks, and restores them once they recover. ([source](https://github.com/benedictking/ccx#readme))
- [Upstream Channel Schedulers](https://awesome-repositories.com/f/artificial-intelligence-ml/upstream-channel-schedulers.md) — Dispatches requests across upstream channels by configured priority, with promotion windows deciding when a preferred channel gets traffic again.
- [Claude Messages Proxies](https://awesome-repositories.com/f/artificial-intelligence-ml/claude-messages-proxies.md) — Forwards Claude Messages calls through one shared address and credential so clients need not connect to each provider separately. ([source](https://github.com/benedictking/ccx#readme))
- [Capability-Based Routing](https://awesome-repositories.com/f/artificial-intelligence-ml/machine-learning/infrastructure/model-evaluation-and-validation/model-capability-assessment/model-capability-queries/capability-based-routing.md) — Filters candidate upstream channels by each model's real context window and output size so requests only reach channels that can handle them.
- [OpenAI Chat Request Proxies](https://awesome-repositories.com/f/artificial-intelligence-ml/openai-chat-request-proxies.md) — Relays OpenAI-style chat and Codex requests through one unified endpoint so existing clients keep working unchanged. ([source](https://github.com/benedictking/ccx#readme))
- [Upstream Channel Configurations](https://awesome-repositories.com/f/artificial-intelligence-ml/upstream-channel-configurations.md) — Adds channels for several upstream service types, each with its own API keys, model name mappings, and custom request parameters. ([source](https://github.com/benedictking/ccx#readme))

### Networking & Communication

- [AI Proxy Gateways](https://awesome-repositories.com/f/networking-communication/ai-proxy-gateways.md) — Provides a unified gateway exposing one endpoint and credential in front of Claude, OpenAI, and Gemini APIs, translating between their protocols.
- [API Protocol Translations](https://awesome-repositories.com/f/networking-communication/cross-chain-messaging-protocols/protocol-translators/messaging-protocol-translation/ai-protocol-translation/api-protocol-translations.md) — Converts requests between Claude, OpenAI, and Gemini API formats and back so clients can use any upstream through one familiar interface.
- [Round-Robin API Key Distributors](https://awesome-repositories.com/f/networking-communication/load-balancing/api-key-load-balancers/round-robin-api-key-distributors.md) — Cycles each channel through several API keys in turn to spread load and avoid rate limits on any single key.

### DevOps & Infrastructure

- [Multi-Provider Key Rotators](https://awesome-repositories.com/f/devops-infrastructure/load-balancing/api-key-rotation/multi-provider-key-rotators.md) — Keeps AI requests flowing by rotating API keys, prioritizing upstream channels, and switching away from failing providers until they recover.
- [API Key Rotation](https://awesome-repositories.com/f/devops-infrastructure/load-balancing/api-key-rotation.md) — Rotates through several API keys per channel, with optional per-channel proxies, custom headers, model allowlists, and route prefixes. ([source](https://github.com/benedictking/ccx#readme))
- [Gateway Deployment](https://awesome-repositories.com/f/devops-infrastructure/remote-management-gateways/gateway-deployment.md) — Runs an AI gateway on your own machine or server as a binary, container, or desktop app, managed through a built-in web console.
- [Single-Binary Deployments](https://awesome-repositories.com/f/devops-infrastructure/single-binary-deployments.md) — Bundles the compiled web console into the backend executable so one process and one port serve both the admin interface and the proxy. ([source](https://github.com/benedictking/ccx#readme))

### Software Engineering & Architecture

- [Node Health-Check Failover](https://awesome-repositories.com/f/software-engineering-architecture/node-health-check-failover.md) — Runs continuous health checks on upstream channels, skips failing ones automatically, and restores them once they recover.
- [Embedded Frontend Bundles](https://awesome-repositories.com/f/software-engineering-architecture/embedded-frontend-bundles.md) — Embeds the compiled web console into the Go backend binary so one process and one port serve both console and proxy.

### System Administration & Monitoring

- [Gateway Admin Consoles](https://awesome-repositories.com/f/system-administration-monitoring/gateway-admin-consoles.md) — Ships a built-in web console and admin API to configure channels, keys, and gateway behavior without editing files. ([source](https://github.com/benedictking/ccx#readme))
- [Gateway Management Consoles](https://awesome-repositories.com/f/system-administration-monitoring/gateway-management-consoles.md) — Ships a built-in browser console for managing channels, API keys, logs, and live per-channel traffic without editing files.
- [Request Traffic Monitors](https://awesome-repositories.com/f/system-administration-monitoring/monitoring-and-observability/observability-platforms/operational-health-alerting/health-monitoring-endpoints/request-traffic-monitors.md) — Tracks per-channel request volume, success rate, and latency in real time to reveal slow or failing upstreams. ([source](https://github.com/benedictking/ccx#readme))
- [AI Traffic Monitors](https://awesome-repositories.com/f/system-administration-monitoring/monitoring-and-observability/observability-platforms/operational-health-alerting/health-monitoring-endpoints/request-traffic-monitors/ai-traffic-monitors.md) — Watches request volume, success rate, and latency for each upstream channel so slow or failing AI providers are spotted quickly.
