Vespa
Open-source engine for search, recommendation and RAG at scale.
An open-source serving engine that combines vector search, lexical search, structured filtering and machine-learned ranking — including multi-vector (ColBERT-style) retrieval — over large datasets in real time. Self-host or use Vespa Cloud.
- Vendor
- Vespa.ai
- Category
- Vector databases
- Pricing
- Open source
- Open source
- Yes
- License
- Apache-2.0
- Platforms
- Cloud, Self-hosted
- Launched
- 2017
- Website
- vespa.ai
Features
- Hybrid search
- Multi-vector retrieval
- ML ranking
Best for
- Research
- Enterprise
- High volume / low cost
Works with
- pplx-embed-v2-late-9b — Perplexity. Perplexity’s large late-interaction embedder for text and visual document retrieval.
More vector databases
- pgvector — Open source. Vector similarity search inside Postgres.
- Pinecone — Pinecone. Fully managed serverless vector database.
- Qdrant — Qdrant. High-performance vector search engine in Rust.
- Chroma — Chroma. The open-source embedding database.
- Weaviate — Weaviate. AI-native open-source vector database.
- Milvus — Zilliz. Vector database for billion-scale workloads.
- Elasticsearch — Elastic. Search engine with dense, sparse and hybrid vector retrieval.
- Redis (vector search) — Redis. In-memory vector search, semantic caching and agent memory.