Choosing a Client

As described in Structured & Unstructured Embeddings, AI workloads come in many forms.

For data science or ephemeral workloads, the Supabase Vecs client gets you started quickly. All you need is a connection string and vecs handles setting up your database to store and query vectors with associated metadata.

You can get your connection string from the Database Settings page in your dashboard. Make sure to check Use connection pooling, then copy the URI. Also, change the URI scheme from postgres to postgresql. vecs uses SQLAlchemy under the hood, which only supports postgresql as a dialect.

For production python applications with version controlled migrations, we recommend adding first class vector support to your toolchain by registering the vector type with your ORM. pgvector provides bindings for the most commonly used SQL drivers/libraries including Django, SQLAlchemy, SQLModel, psycopg, asyncpg and Peewee.