Supabase Vector
Freshby Supabase
Vector similarity search and embedding storage via Postgres extensionAI engineers building RAG pipelines or semantic search who want embeddings co-located with relational data, avoiding a separate vector databaseFree tier
pgvector-powered embedding storage inside the project Postgres database, supporting HNSW and IVFFlat indexes up to 16,000 dimensions with SQL-native nearest-neighbor search.
Features
Row Level Security (RLS)all plansAuto-generated REST API (PostgREST)all plansAuto-generated GraphQL API (pg_graphql)all plansDatabase Branching (git-synced)paid onlyPoint-in-Time Recoverypaid onlypgvector Embeddings (HNSW/IVFFlat)all plansTOTP Multi-Factor Authenticationall plansPasskeys / WebAuthnall plansEnterprise SSO (SAML 2.0)enterprise onlyOn-the-fly Image Transformationspaid onlyRead Replicas (multi-region)paid onlySOC2 & ISO 27001 Compliancepaid onlyHIPAA Add-onpaid only28-day Log Retentionpaid onlyDatabase Triggers for Edge Functionsall plans40+ Postgres Extensions (one-click)all plansCompute Autoscalingpaid only
More from Supabase
- Supabase Auth
Identity and user management with database-level authorization via Row Level Security
- Supabase Database
Managed relational database with schema-driven API generation and connection pooling
- Supabase Edge Functions
Serverless edge compute with database trigger support and npm ecosystem compatibility
- Supabase Realtime
Change data capture and WebSocket pub/sub multiplexed over Postgres
- Supabase Storage
Object storage with CDN delivery, RLS-based access control, and server-side image processing