Ollama
Python library for calling Ollama models from your own Python code.ollama/ollama-python
- Stars
- 10.7k
- Forks
- 1.2k
- Licence
- MIT
- Language
- Python
- Last update
- 20 days ago
- Type
- Library
What it is
This is a Python library that lets Python programs send chat, generation, and embedding requests to a model running through Ollama. You need Ollama installed and running, and a model pulled, before the library can do anything. It wraps Ollama's REST API so calls look like ordinary Python functions, with support for streaming and async use.
- Ask a model a question with chat() and print the reply
- Stream a long answer token by token into a terminal or web page
- Create embeddings for a list of texts with ollama.embed()
- Run cloud-hosted models through the Client with an API key set in OLLAMA_API_KEY
Great for
- Python developers adding a local or cloud language model to a script or app
- People who want to call Ollama from Python without writing raw HTTP requests
- Projects that need streaming responses or async requests
- Developers building embedding or batch text-processing pipelines
Not ideal for
- People who do not write Python: the README only covers Python usage
- Users who want a chat interface without coding, since this is a library rather than an app
- Projects needing Python older than 3.8, which the README says is the minimum
How to get it
- 1
Install with a package manager
You'll need: Python 3.8+ · Ollama installed and running
Install the library
pip install ollama - 2
Run it
- Install Ollama from ollama.com/download and make sure it is running
- Pull a model, for example: ollama pull gemma4
- Install the library with pip install ollama
- Run the chat example from the README with model='gemma4'
- 3
Get the most out of it
- Use stream=True to print partial output as it arrives
- Use AsyncClient for asynchronous apps
- Pass custom headers or a different host to Client to point at another server
- Use the Client options to pass extra keyword arguments to the underlying httpx.Client
- Catch ollama.ResponseError and pull a missing model on a 404 status
Downloads
Strengths
- Simple, documented function and client interfaces with examples for sync and async use
- Covers the main Ollama REST API operations, including chat, pull, embed, and model management
- Released under the MIT license with a visible release history
Trade-offs
- Requires a separate Ollama install and a pulled model; the library alone does nothing
- The README is truncated in this record, so some details may be missing from this entry
- Cloud features depend on Ollama account sign-in and API keys
How it compares
| Project | Stars | Type | Level | Updated | Best for |
|---|---|---|---|---|---|
| ollama-python (this) | 10.7k | Library | beginner | 20 days ago | Python developers adding a local or cloud language model to a script or app |
| ollama | 186k | App | beginner | 4 days ago | Developers who want a local model API for their own apps |
| deepwiki-open | 18.3k | App | intermediate | 8 days ago | Developers trying to understand a large or unfamiliar codebase |
| pr-agent | 13.9k | Dev tool | intermediate | 5 days ago | Teams that want automated first-pass reviews on pull requests |
| LEANN | 13.8k | Library | intermediate | 4 days ago | Developers building a private RAG prototype on a personal machine |
Stats synced from GitHub. The assessment above was drafted with AI from the project's README on 2026-10-29 and may miss recent changes — check the README before relying on a command. MIT licence: free to use, modify and distribute, including commercially, as long as you keep the copyright notice.