Serve models
Tutorial 2 covers basic single-model serving. These guides cover configuration beyond a single local server, including OpenAI-compatible /v1 clients and Anthropic-compatible /v1/messages clients.
Choose by task
Section titled “Choose by task”| If you need to… | Start here |
|---|---|
| Change host, port, CORS, request limits, or authentication | HTTP server configuration |
| Serve more than one model from one process | Running multiple models |
| Use the browser chat interface | Using the web UI |
| Use OpenAI-compatible clients | OpenAI-compatible APIs |
| Use the newer OpenAI Responses endpoint | OpenAI Responses API |
| Use Anthropic-compatible clients | Anthropic Messages API |
| Use Codex or Claude Code with a local server | Use Codex and Claude Code |
For operational concerns (reverse proxy, Docker, health checks, TLS), see the deployment guides.