Best Vector Databases Software & Tools

Compare vector databases on user reviews: recall versus latency, index build times, filtering support, scaling behavior and cost per million vectors.

10 products

Amazon OpenSearch Service is an AWS-managed search, log analytics and vector database, run as sized clusters or as usage-billed OpenSearch Serverless.

Features

Has Approximate nearest-neighbor index
Has Filtered vector search on metadata
Has Hybrid keyword and vector search
Has Choice of similarity metrics
Has Quantized index compression
Missing Multi-tenant namespaces per index
Has Integrated embedding models
Missing Built-in reranking stage
Missing Multiple named vectors per record
Has Live upserts visible in seconds
Has Sharded vector collections
Has Replicated vector shards
Has Disk-backed index beyond RAM
Has Collection snapshots and restore
Has Role-based keys per collection
Has Bulk import from Parquet files
Has Serverless usage-based pricing
Has Open source self-hosted engine
Chroma

by Chroma

No reviews yet

Chroma is an open-source vector database for AI retrieval, licensed Apache 2.0 to self-host or run as Chroma Cloud, a serverless service billed by usage.

Features

Has Approximate nearest-neighbor index
Has Filtered vector search on metadata
Has Hybrid keyword and vector search
Has Choice of similarity metrics
Missing Quantized index compression
Missing Multi-tenant namespaces per index
Has Integrated embedding models
Missing Built-in reranking stage
Missing Multiple named vectors per record
Missing Live upserts visible in seconds
Missing Sharded vector collections
Missing Replicated vector shards
Missing Disk-backed index beyond RAM
Has Collection snapshots and restore
Missing Role-based keys per collection
Missing Bulk import from Parquet files
Has Serverless usage-based pricing
Has Open source self-hosted engine
Elasticsearch

by Elastic

No reviews yet

Elasticsearch is Elastic's open source search engine and vector database for developers who need keyword, vector and hybrid search, self-hosted or managed.

Features

Has Approximate nearest-neighbor index
Has Filtered vector search on metadata
Has Hybrid keyword and vector search
Has Choice of similarity metrics
Has Quantized index compression
Missing Multi-tenant namespaces per index
Has Integrated embedding models
Has Built-in reranking stage
Has Multiple named vectors per record
Has Live upserts visible in seconds
Has Sharded vector collections
Has Replicated vector shards
Has Disk-backed index beyond RAM
Has Collection snapshots and restore
Has Role-based keys per collection
Missing Bulk import from Parquet files
Has Serverless usage-based pricing
Has Open source self-hosted engine
LanceDB

by LanceDB

No reviews yet

LanceDB is an open source vector database and multimodal lakehouse for AI teams, run as an embedded library or as an Enterprise cluster in a private cloud.

Features

Has Approximate nearest-neighbor index
Has Filtered vector search on metadata
Has Hybrid keyword and vector search
Has Choice of similarity metrics
Has Quantized index compression
Missing Multi-tenant namespaces per index
Has Integrated embedding models
Has Built-in reranking stage
Has Multiple named vectors per record
Has Live upserts visible in seconds
Missing Sharded vector collections
Has Replicated vector shards
Has Disk-backed index beyond RAM
Has Collection snapshots and restore
Missing Role-based keys per collection
Has Bulk import from Parquet files
Missing Serverless usage-based pricing
Has Open source self-hosted engine
Pinecone

by Pinecone

No reviews yet

Pinecone is a managed, serverless vector database for developers building semantic search, RAG and AI agents, with hosted embedding and reranking models.

Features

Has Approximate nearest-neighbor index
Has Filtered vector search on metadata
Has Hybrid keyword and vector search
Has Choice of similarity metrics
Missing Quantized index compression
Has Multi-tenant namespaces per index
Has Integrated embedding models
Has Built-in reranking stage
Has Multiple named vectors per record
Has Live upserts visible in seconds
Has Sharded vector collections
Has Replicated vector shards
Missing Disk-backed index beyond RAM
Has Collection snapshots and restore
Missing Role-based keys per collection
Has Bulk import from Parquet files
Has Serverless usage-based pricing
Missing Open source self-hosted engine
Qdrant

by Qdrant

No reviews yet

Qdrant is an open source vector database with managed, hybrid and private cloud editions, for teams building RAG, semantic search and agent memory.

Features

Has Approximate nearest-neighbor index
Has Filtered vector search on metadata
Has Hybrid keyword and vector search
Has Choice of similarity metrics
Has Quantized index compression
Has Multi-tenant namespaces per index
Has Integrated embedding models
Has Built-in reranking stage
Has Multiple named vectors per record
Has Live upserts visible in seconds
Has Sharded vector collections
Has Replicated vector shards
Has Disk-backed index beyond RAM
Has Collection snapshots and restore
Has Role-based keys per collection
Missing Bulk import from Parquet files
Missing Serverless usage-based pricing
Has Open source self-hosted engine
turbopuffer

by turbopuffer

No reviews yet

turbopuffer is a serverless vector and full-text search database on object storage, for teams running AI search over large multi-tenant data. Billed by usage.

Features

Has Approximate nearest-neighbor index
Has Filtered vector search on metadata
Has Hybrid keyword and vector search
Missing Choice of similarity metrics
Missing Quantized index compression
Has Multi-tenant namespaces per index
Has Integrated embedding models
Has Built-in reranking stage
Has Multiple named vectors per record
Has Live upserts visible in seconds
Has Sharded vector collections
Has Replicated vector shards
Has Disk-backed index beyond RAM
Has Collection snapshots and restore
Missing Role-based keys per collection
Missing Bulk import from Parquet files
Has Serverless usage-based pricing
Missing Open source self-hosted engine
Vespa

by Vespa

No reviews yet

Vespa is an open-source AI search platform and vector database from Norway's Vespa.ai, for RAG, search and recommendation, self-hosted or on Vespa Cloud.

Features

Has Approximate nearest-neighbor index
Has Filtered vector search on metadata
Has Hybrid keyword and vector search
Has Choice of similarity metrics
Has Quantized index compression
Has Multi-tenant namespaces per index
Has Integrated embedding models
Has Built-in reranking stage
Has Multiple named vectors per record
Has Live upserts visible in seconds
Has Sharded vector collections
Has Replicated vector shards
Has Disk-backed index beyond RAM
Has Collection snapshots and restore
Missing Role-based keys per collection
Missing Bulk import from Parquet files
Missing Serverless usage-based pricing
Has Open source self-hosted engine
Weaviate

by Weaviate

No reviews yet

Weaviate is an open-source vector database for RAG, hybrid search and agent memory, run self-hosted or managed on Weaviate Cloud with usage-based plans.

Features

Has Approximate nearest-neighbor index
Has Filtered vector search on metadata
Has Hybrid keyword and vector search
Has Choice of similarity metrics
Has Quantized index compression
Has Multi-tenant namespaces per index
Has Integrated embedding models
Has Built-in reranking stage
Has Multiple named vectors per record
Has Live upserts visible in seconds
Has Sharded vector collections
Has Replicated vector shards
Has Disk-backed index beyond RAM
Has Collection snapshots and restore
Has Role-based keys per collection
Missing Bulk import from Parquet files
Missing Serverless usage-based pricing
Has Open source self-hosted engine
Zilliz Cloud

by Zilliz

No reviews yet

Zilliz Cloud is a managed vector database built on open-source Milvus, for developers adding RAG, semantic search or recommendations to their apps.

Features

Has Approximate nearest-neighbor index
Has Filtered vector search on metadata
Has Hybrid keyword and vector search
Has Choice of similarity metrics
Missing Quantized index compression
Has Multi-tenant namespaces per index
Has Integrated embedding models
Has Built-in reranking stage
Has Multiple named vectors per record
Missing Live upserts visible in seconds
Missing Sharded vector collections
Has Replicated vector shards
Has Disk-backed index beyond RAM
Has Collection snapshots and restore
Has Role-based keys per collection
Has Bulk import from Parquet files
Has Serverless usage-based pricing
Has Open source self-hosted engine
0 selected