AI GitHub
turbopuffer Python client for hosted retrieval
Connect Python applications to turbopuffer's hosted search API.
Key Takeaways
- The repository is the official client for managing and querying data in turbopuffer namespaces.
- Use the service endpoint and credentials for a test namespace.
- Use a disposable namespace and a few synthetic records.
- This is a client for an external service.
How it works
The repository is the official client for managing and querying data in turbopuffer namespaces. The hosted service supplies storage and retrieval, and the client handles request construction and authentication.
What you need
Use the service endpoint and credentials for a test namespace. Keep request timeouts, retry behavior and access scope explicit in the Python integration.
What to try first
Use a disposable namespace and a few synthetic records. Verify vector queries, metadata filters and deletion before connecting an ingestion job.
Test a query before and after updating one record and deleting another. Confirm how quickly the namespace reflects changes your application needs to see.
Where to look for failures
This is a client for an external service. Check data handling, credentials and service limits with the provider for the deployment you intend to use.
Related Franklin pages
Separate the search interface from where the data lives. A small import, query, update and delete test reveals important differences between client, embedded and server deployments.
- Query embeddings and associated records in tables based on the Lance format. LanceDB for retrieval over columnar data
- Search embeddings alongside structured metadata in a dedicated database. Qdrant for filtered vector retrieval
For the role of the index in nearest-neighbor search, see Franklin’s glossary. Vector indexes
Project documentation
Try it for your workflow
Explore the project’s documentation for setup and usage, or save this page to a collection for later.
Read the project documentation ↗