vecparity

Live, quality-verified migration between vector databases.

Every team on a managed vector database eventually wants to switch: Pinecone got expensive, you want to self-host, a competitor is faster for your workload. Existing tools stop at "the data arrived": vector-io/VDF does batch export → import → re-embed across backends, and vendor migration guides only cover moving into their own product. None of them check whether retrieval quality survived the move.

vecparity does two things nothing else does together:

• Migrates incrementally, without downtime. A SyncEngine that polls a source adapter's list_changed_since cursor and replays changes to the target, so a migration can run alongside a live app.
• Verifies retrieval parity before you cut over. verify_parity() runs a query set against both the source and target and reports recall@k, result overlap, and score drift, producing a pass/fail ParityReport you can gate a deploy on.

Where to go next

• Installation: install with the backends you need
• Quick Start: migrate and verify from the CLI or Python
• Backends: connection details and env vars per backend
• Design Principles: why the adapter surface is deliberately small
• API Reference: full module documentation