Typesense
by Typesense · Open-source in-memory search engine with a REST API
45.2 — BenchRank score out of 100
Screenshots of Typesense
Homepage Pricing page
Overview
Typesense is an open-source, in-memory search engine with a REST API, aimed at search-as-you-type experiences. It supports typo tolerance, filtering and faceting, geo search, grouping, synonyms, merchandising, vector and semantic search, and federated search across sources. Run it yourself via Docker, binaries or DEB/RPM packages, or use Typesense Cloud for managed clusters.
- Best for
- Developers adding typo-tolerant instant search to an app, self-hosted or on managed clusters.
- Pricing
- Self-hosting is free and open source; Typesense Cloud bills hourly from $0.03/hr (stated as about $21.60/month) plus $0.09/GB bandwidth out, with the first 720 hours and first 10 GB free.
- Runs on
- LinuxSelf-hosted
Strengths and trade-offs
Strengths
- Open source; runs via Docker, native binaries or DEB/RPM packages
- Typo tolerance, faceting, geo, vector and semantic search built in
- Official clients for JS, PHP, Ruby, Go, Python, Java, Swift and more
- Cloud clusters have no limits on records or operations
Trade-offs
- In-memory design: cloud cost tracks the RAM you provision, not record count
- You size RAM, vCPUs, HA and regions yourself; no fixed plans
- High availability and Search Delivery Network cost extra on top of the cluster
- SOC 2 reports, HIPAA BAAs and security questionnaires need a paid support plan
How this score is made up
Each dimension is scored out of 100 and combined into the headline score using fixed weights.
- MCP support
- 0 out of 100
- API quality
- 50 out of 100
- Documentation
- 60 out of 100
- Agent friendliness
- 78 out of 100
- Pricing transparency
- 100 out of 100
- Changelog
- 0 out of 100
- Marketing site structure
- 50 out of 100
- Operational trust
- 15 out of 100
Measured, but not part of the score
Useful to know, but not a mark for or against the product — so these do not affect the ranking.
- Openness
- 40 out of 100
This doesn’t look right — report a problem with Typesense’s score
Alternatives in Vector & Retrieval
Ranked 1
78.8 — BenchRank score out of 100Meilisearch
Meilisearch · Open-source search engine with full-text, semantic and hybrid retrieval
Best for: Development teams adding full-text, semantic or hybrid search to an app, self-hosted or managed
Ranked 2
69.5 — BenchRank score out of 100Redis
Redis · In-memory data platform for caching, vector search and agent context
Best for: Teams needing a low-latency store for caching, vector search and agent context, hosted or self-run
Ranked 3
66.1 — BenchRank score out of 100SWIRL
Swirl Search · Federated enterprise search and MCP layer that leaves data in place
Best for: Regulated enterprises wanting AI search over existing systems without copying data to a new index
