rocketkind
AI

Ollama

Ollama runs open-weight models on hardware you control and exposes them behind an OpenAI-compatible HTTP API.

Authentication

  • API key
  • No authentication

Actions

  • Chat CompletionsAsk a model running on your own Ollama server for one complete answer, over its OpenAI-compatible endpoint (POST /v1/chat/completions).
  • Chat Completions (Streaming)Ask a model running on your own Ollama server for an answer and RETURN it as a live stream of Server-Sent-Events chunks, one per response delta, so a watching caller sees tokens as they are produced instead of waiting for the whole reply.
  • CompletionsOpenAI-compatible POST /v1/completions (legacy text completion) against an Ollama server.
  • EmbeddingsOpenAI-compatible POST /v1/embeddings against an Ollama server.
  • List local models (native)Native GET /api/tags probe using connection metadata.apiBaseUrl (the trailing /v1 compatibility segment is stripped).
  • List modelsOpenAI-compatible GET /models probe using connection metadata.apiBaseUrl.
  • Retrieve modelOpenAI-compatible GET /v1/models/{model} for one model’s metadata from an Ollama server.

Product names, logos and brands are the property of their respective owners and are used here for identification only. Their use does not imply endorsement by, or affiliation with, those owners.