Ollama¶

Use the ollama provider to call a local Ollama server or Ollama Cloud. It supports chat and embeddings. Chat requests default to Chat Completions; stateless Responses requires Ollama 0.13.3 or later.

Get started¶

Install Republic using the quickstart, start Ollama, and replace MODEL_ID with a pulled model:

import asyncio

import republic


async def main():
    model = republic.get_model("ollama:MODEL_ID")
    response = await model.chat("Say hello in one sentence.")
    print(response.text)


asyncio.run(main())

Authentication¶

A local server needs no key, and Republic sends no Authorization header unless REPUBLIC_OLLAMA_API_KEY or api_key= is set. For Ollama Cloud, set REPUBLIC_OLLAMA_API_BASE=https://ollama.com/v1 and an API key from your Ollama account.

Model support¶

The default endpoint is http://localhost:11434/v1. Chat requests send max_tokens. Select api_format="responses" for stateless Responses. Use get_embedding_model() for embedding models.

See configuration for endpoint overrides and request options.