ollama/ollama-python

Project overview

It bridges Python applications and Ollama models while supporting both local execution and optional cloud offloading, giving developers flexibility in how they run large language models.

Project type

Model Runtime

Deployment

Refer to project documentation

License

MIT

Best for

• Developers building Python 3.8+ applications who need a straightforward library to interact with Ollama models.
• AI engineers who want to run models locally or offload larger models to Ollama's cloud.

Key capabilities

• Enables chat interactions with models, including streaming responses.
• Generates responses from a model based on a prompt.
• Provides an AsyncClient for making asynchronous requests.
• Lists, shows, creates, copies, deletes, pulls, and pushes models.

Limitations and risks

• Requires a running instance of Ollama to function. The library cannot operate standalone without an accessible Ollama server.
• Models such as gemma3 must be pulled via Ollama before they can be used through the library.

Getting started

• Installable via pip with a single command: run `pip install ollama`.
• Ensure Ollama is installed and running. A model such as gemma3 must be pulled via Ollama before use.

Evidence and sources

• README: The Ollama Python library provides the easiest way to integrate Python 3.8+ projects with Ollama.
• README: - Ollama should be installed and running - Pull a model to use with the library: ollama pull <model> e.g. ollama pull gemma3
• README: response: ChatResponse = chat(model='gemma3', messages=[ { 'role': 'user', 'content': 'Why is the sky blue?', }, ]) print(response['message']['content'])
• README: Response streaming can be enabled by setting `stream=True`.
• README: ollama.generate(model='gemma3', prompt='Why is the sky blue?')

Reviewed at: 2026-07-18T03:03:45Z