Sends a single chat message to the running Ollama server over HTTP, with the conversation passed as a messages array exactly like OpenAI's API shape. Add "stream":false to get the full reply in one JSON response instead of tokens as they are generated. This is how other programs talk to Ollama; `ollama run llama3.2` is the interactive shortcut.
Looking for more? Search all 7,657 commands — works offline, in English or Spanish, and fixes typos.