vllm-project/vllm
API
vLLM exposes its inference engine through several network APIs and a CLI. This section describes them.
- OpenAI-compatible server — the primary HTTP API
- Anthropic, gRPC, MCP, SageMaker frontends — alternative shells over the same engine
Built by Factory AutoWiki from public repository content. It is a generated preview for codebase exploration, not source-maintained documentation.