Category
Vector Database APIs pricing
Vector database APIs power semantic search, RAG, and AI memory by storing and querying high-dimensional embeddings. The category splits into purpose-built managed services (Pinecone, Weaviate, Qdrant, Zilliz/Milvus, Chroma, Turbopuffer) and vector capabilities bolted onto general databases (pgvector via Supabase, MongoDB Atlas). Compare on documentation/DX quality, reliability (published SLAs and proven scale), SDK/ecosystem breadth, and how fast a developer or AI agent can self-serve a working API key. Pricing models diverge sharply: usage-based per read/write/storage units (Pinecone, Turbopuffer, Chroma), resource-based vCPU/RAM (Qdrant), compute-unit (Zilliz), or bundled into a broader DB plan (Supabase, MongoDB).
| API | Billed by | Free tier | Verified |
|---|---|---|---|
 Pinecone Vector Database APIs | Per read/write unit + storage | Free tier or trial | Jun 27, 2026 |
 MongoDB Atlas Vector Search Vector Database APIs | Atlas cluster consumption-based | Free tier or trial | Jun 27, 2026 |
 Zilliz Cloud (Milvus) Vector Database APIs | Per compute-unit-hour + storage | Free tier or trial | Jun 27, 2026 |
 Supabase Vector (pgvector) Vector Database APIs | Bundled in Postgres plan (storage-based) | Free tier or trial | Jun 27, 2026 |
 Qdrant Cloud Vector Database APIs | Resource-based (vCPU/RAM/disk) | Free tier or trial | Jun 27, 2026 |
 Weaviate Cloud Vector Database APIs | Usage-based (dimensions + storage) | Free tier or trial | Jun 27, 2026 |
 Turbopuffer Vector Database APIs | Per GB queried/written + storage | No free tier | Jun 27, 2026 |