Any endpoint routes through the ChatGPT, Claude, Grok, and Kimi subscriptions you already pay for. No per-token pricing, no markup. Works in your app, IDE, and CLI.
- Stop paying twice. Bring the subscriptions, API keys, and custom endpoints you already use into one consistent surface.
- Stop rewriting integrations. Chat, responses, embeddings, images, and audio sit behind OpenAI- and Anthropic-compatible APIs.
- Build resilient chains. Route across providers and move to the next model when one is down, rate-limited, or over context.
- Keep credentials yours. Provider keys are encrypted under recovery material only you hold; subscription providers run locally through official vendor CLIs.
- Protocol — runtime-validated contracts shared by OpenLLM clients.
- Wire — pure transforms across Anthropic, OpenAI, and Responses wire formats.
- Tunnel — sans-I/O binary mux for consumer ↔ relay ↔ daemon streams.
- CLI — the single
openllmbinary: client launcher, MCP gateway with semantic search and memory, and the local daemon (openllm daemon) for subscription-backed provider CLIs.
BUILD WITH THE MODELS YOU ALREADY HAVE
