Welcome to Weaviate Python Client’s documentation!
The Weaviate Python client makes it easy to work with Weaviate, an AI Native vector database. The client is tested for Python 3.8 and higher.
The current major version is v4 while the older v3 client is deprecated and should be avoided. You can find further documentation for both versions:
Python Client v3 (deprecated)
Note
Follow the Weaviate Quickstart guide to get up and running quickly.
Installation
You can install the Weaviate Python client using pip:
pip install -U weaviate-client
See the installation section in the main Weaviate documentation for more details.
Optional: Weaviate Agents
This client supports optional integration with Weaviate Agents, which are agentic pre-built services designed to simplify common LLM-related tasks like querying, transformation, and personalization using your Weaviate data.
To install the client with agent support, use the [agents] extra:
pip install -U "weaviate-client[agents]"
For detailed guides and usage examples for the different agents, please refer to the main Weaviate Agents documentation.
Note
Weaviate Agents require a connection to a Weaviate Cloud (WCD) instance, local instances are not supported.
Client API reference
Explore the detailed API documentation:
Support
Use the
weaviatetag on StackOverflow for questions.For bugs or problems, submit a GitHub issue.