API
Point clients at orouta the same way you would point them at Ollama.
Ollama
Section titled “Ollama”POST /api/chat, /api/generate, /api/embed, /api/pull, and the rest of the Ollama HTTP API are forwarded to the host that listed model or name in /api/tags. Streaming is passed through.
POST /api/pull downloads a model onto the host that would serve it. Add ?host=<id> to pick a different upstream: POST /api/pull?host=desk with {"model":"llama3"} pulls onto desk and streams progress back like Ollama does.
Unknown names return 404 {"error":"unknown model"}. There is no download map in config: a host is chosen by what its /api/tags lists, and new weights arrive via POST /api/pull (above) or a manual ollama pull on that machine.
OpenAI
Section titled “OpenAI”Ollama already serves /v1/chat/completions, /v1/completions, /v1/embeddings, and /v1/models. orouta forwards those paths. Set base_url to http://127.0.0.1:11434/v1 and the API key to a value from auth.keys.
Anthropic Messages
Section titled “Anthropic Messages”POST /v1/messages is translated to Ollama POST /api/chat. Text only. max_tokens becomes options.num_predict. Tools, images, and non-text blocks return 400.
curl -H "x-api-key: sk-orouta-alice" \ -H "Content-Type: application/json" \ -d '{"model":"llama3","max_tokens":64,"messages":[{"role":"user","content":"hi"}]}' \ http://127.0.0.1:11434/v1/messagesThe JSON model field still picks the Ollama host. There is no Anthropic upstream.