Vector Store | PrivateGPT

PrivateGPT uses Qdrant as the default vector store. By default it runs in local disk mode — no separate Qdrant server needed.

Local disk mode is not recommended for production. Use a remote Qdrant server for any persistent or multi-instance deployment.

1 qdrant:
2   path: local_data/private_gpt/qdrant

Remote Qdrant

To connect to an existing Qdrant server instead of the local disk instance:

1 qdrant:
2   url: http://localhost:6333
3   api_key: your-api-key   # only needed for Qdrant Cloud

If you switch embedding models, the vector dimensions change. Re-ingest your documents after changing the model and update embed_dim to match:

1 vectorstore:
2   embed_dim: 1024   # must match the output dimension of your embedding model

Run make wipe before re-ingesting to avoid a dimension mismatch error.

1	qdrant:
2	url: http://localhost:6333
3	api_key: your-api-key # only needed for Qdrant Cloud

1	vectorstore:
2	embed_dim: 1024 # must match the output dimension of your embedding model