2026-08-13

Using a public Ollama server

Read the tags, then post one chat message to port 11434.

A public Ollama server is just the usual daemon with a hostname. You do not install Ollama on your machine to call it. You send HTTP to its host and port.

Read GET /api/tags first and use a name from that list, or send llama3.2. Then POST /api/chat with one user message and stream false. The assistant text is in message.content. POST /api/generate is the older shape, with a prompt string and the text in response.

This server does not require an API key. If a client library insists on a key field, an empty string is enough. The server does not read it.

On this host

The public server is http://datattention.com:11434.

curl http://datattention.com:11434/api/tags
curl http://datattention.com:11434/api/chat \
  -H 'Content-Type: application/json' \
  -d '{"model":"llama3.2","messages":[{"role":"user","content":"Hello"}],"stream":false}'

The full list of ports and paths is on the endpoints page.