Typesense
activeby Typesense, Inc. · since 2017
In-memory, typo-tolerant search engine in C++ with vector, hybrid and conversational search.
Overview
Typesense is an open-source search engine that keeps its index in RAM for consistently low-latency, typo-tolerant queries and pitches itself as a self-hostable Algolia alternative. Beyond keyword search it supports vector and hybrid search with built-in embedding models, joins between collections, geo search and a conversational (RAG) endpoint. Clustering uses Raft for high availability, but every node holds the full dataset, so capacity is bounded by the RAM of one machine. It is GPL-3.0 licensed, with Typesense Cloud as the managed option.
Strengths & trade-offs
Strengths
- +Predictable millisecond latency because the index lives in RAM
- +Typo tolerance, synonyms, curation and faceting with little tuning
- +Built-in embedding models for semantic and hybrid search
- +Simple Raft clustering for high availability
- +No per-search pricing on the managed cloud
Trade-offs
- −Dataset must fit in RAM on every node
- −GPL-3.0 can be a hurdle for embedding in proprietary distributions
- −No sharding, so very large catalogs need app-level splitting
- −Smaller ecosystem than Elasticsearch
Great for
- +E-commerce and marketplace search
- +Docs and website search
- +Geo search for listings
- +Replacing Algolia to cut per-request costs
Think twice for
- −Datasets far larger than RAM
- −Log analytics
- −Analytical aggregations
Capabilities
Native or official first-party support · 9 of 25 tracked
Scale & consistency
- Comfortable data size
- Small · ≤ 100 GB
- Horizontal scaling
- None
- Multi-region
- Read replicas
- Transactions
- Limited ACID
- Consistency
- Strong
- Schema
- Flexible
- Architecture
- Primary replica, In memory, Client server
- Query languages
- Typesense search API (JSON)
Raft-replicated clusters for HA; each node stores the whole index in RAM. Typesense Cloud's Search Delivery Network places replicas in multiple regions.
Deployment
Ecosystem
Alternatives to Typesense
Developer-friendly Rust search engine with typo tolerance and built-in hybrid semantic search.
typesense vs meilisearch →Hosted search-as-a-service with fast typo-tolerant search, merchandising and AI ranking.
typesense vs algolia →Lucene-based distributed search and analytics engine for text, vectors, logs and metrics.
typesense vs elasticsearch →Tiny TypeScript full-text, vector and hybrid search engine for browsers, servers and the edge.
typesense vs orama →Fast SQL-first search database, forked from Sphinx, with columnar storage and vector KNN.
typesense vs manticore-search →Big-data serving engine for search, recommendation and RAG with tensors and ML ranking.
typesense vs vespa →Typesense FAQ
What is Typesense?
Typesense is a search engine database by Typesense, Inc., first released in 2017. Typesense is an open-source search engine that keeps its index in RAM for consistently low-latency, typo-tolerant queries and pitches itself as a self-hostable Algolia alternative. Beyond keyword search it supports vector and hybrid search with built-in embedding models, joins between collections, geo search and a conversational (RAG) endpoint. Clustering uses Raft for high availability, but every node holds the full dataset, so capacity is bounded by the RAM of one machine. It is GPL-3.0 licensed, with Typesense Cloud as the managed option.
Is Typesense open source?
Yes. Typesense is open source. License: GNU GPL v3.
Does Typesense support ACID transactions?
Partly. Typesense offers limited ACID guarantees (for example, single-record or single-partition transactions), with strong consistency.
Does Typesense scale horizontally?
Typesense does not scale horizontally; it runs on a single node, so you scale it up rather than out. It is comfortable at small scale (≤ 100 GB). It supports cross-region read replicas.
How can I run Typesense?
You can self-host it and use the vendor's managed cloud. Managed offerings include Typesense Cloud.
What is Typesense best for?
Typesense is a strong fit for E-commerce and marketplace search; docs and website search; and geo search for listings. Think twice for datasets far larger than RAM.
What are the best alternatives to Typesense?
Common alternatives to Typesense include Meilisearch, Algolia, Elasticsearch, Orama and Manticore Search. The right choice depends on your workload, scale and hosting; WhichDB's smart match ranks them against your requirements.
How popular is Typesense?
Typesense ranks #95 of 844 databases on the WhichDB Score (28.8/100), and #9 among search engine databases. The score combines public adoption signals such as package downloads, GitHub activity, job posts and developer discussion.