Search & Vector Databases brand directory

Meilisearch is a unified search and AI retrieval platform offering an open-source search engine API and a managed cloud service. It provides hybrid full-text and semantic search, vector storage, typo tolerance, geosearch, faceted filtering, and multi-tenancy out of the box — with near-zero configuration required to get started.

Elastic builds and operates the Elasticsearch platform—the world's most widely deployed search and analytics engine—along with the Elastic Stack (ELK). The platform provides distributed full-text search, vector database capabilities, hybrid search (BM25 + dense/sparse vectors via ELSER), real-time analytics, log management, observability, and AI-driven security. It is available as a fully managed serverless cloud service, a hosted cloud deployment, or self-managed on any infrastructure.

Pinecone is a purpose-built, fully managed serverless vector database designed for production AI applications. It provides high-performance approximate nearest neighbor (ANN) search over dense and sparse vector embeddings, supporting hybrid retrieval that combines semantic understanding with exact keyword matching. The platform offers integrated embedding and reranking inference, real-time indexing, metadata filtering, and namespace-based multitenancy. Additional products include Pinecone Assistant (a managed RAG Q&A API) and Dedicated Read Nodes for predictable high-throughput workloads. Pinecone runs on AWS, Azure, and GCP with enterprise BYOC deployment and is compliant with SOC 2, GDPR, ISO 27001, and HIPAA.

Qdrant is a high-performance, open-source vector database and search engine written in Rust, offering native hybrid (dense + sparse) retrieval, one-stage filtered HNSW indexing, multi-vector support, and advanced quantization. It serves as the retrieval backbone for RAG pipelines, AI agents, semantic search, and recommendation systems, and is deployable as open-source, fully managed cloud, hybrid cloud, private cloud, or edge.

Algolia is an AI-powered search and discovery platform delivered as a fully managed SaaS. Its core offering combines keyword and semantic (vector) retrieval in a single API via NeuralSearch, enabling millisecond-latency search with built-in typo tolerance, faceting, and relevance rules. The platform extends into AI Recommendations, behavioral personalization, category-page Browse, a no-code Merchandising Studio, Generative Experiences for RAG-based conversational answers, Ask AI for in-search-bar answers, and an Agent Studio for agentic AI workflow deployment. An Intelligent Data Kit handles data ingestion, transformation, and enrichment. Algolia targets both developers (REST API, InstantSearch UI libraries, 70+ language SDKs) and non-technical business teams (visual editor, analytics dashboards, merchandising controls), positioning itself as a full-stack retrieval layer for e-commerce, SaaS, media, and enterprise applications.

Typesense is a lightning-fast, open-source search engine built in C++ that combines traditional keyword search with modern vector and semantic search capabilities in a single, easy-to-operate package. It is available as a self-hosted binary or as Typesense Cloud, a managed SaaS service. Core strengths include automatic typo tolerance, sub-50ms query latency, hybrid search, built-in RAG/conversational search, geo search, faceted navigation, and a developer-friendly REST API with clients across 10+ languages.

Weaviate is an open-source, cloud-native vector database designed for AI engineers building production-grade search, RAG, and agentic AI applications. It combines dense vector search, BM25 keyword search, and hybrid fusion in one system, with built-in support for multimodal data types, pluggable vectorizer modules, multi-tenancy, horizontal scaling, and enterprise compliance. A managed Weaviate Cloud service eliminates operational overhead, while the open-source version enables full self-hosted control. In 2024–2025, Weaviate expanded beyond a pure vector database into an AI application platform with the introduction of Weaviate Agents (Query, Transformation, and Personalization agents) and a native embedding service.

Zilliz provides open-source (Milvus) and fully managed (Zilliz Cloud) vector database technologies purpose-built for AI applications requiring high-performance embedding similarity search at billion-vector scale. Its proprietary Cardinal Search Engine powers Zilliz Cloud's performance advantages, while its AutoIndex feature removes manual tuning. The platform supports serverless, dedicated, and BYOC deployments across major clouds, integrates with leading AI frameworks, and delivers enterprise-grade security, compliance, and observability tooling.

Vespa.ai is an AI search platform that combines vector search, full-text search, structured data filtering, and machine-learned ranking into a single distributed serving engine. Originally built within Yahoo and open-sourced in 2017, it is designed for applications that must query, rank, and make inferences over billions of continuously changing data items at sub-100ms latencies and thousands of queries per second. It is offered as open-source software and as Vespa Cloud, a managed cloud service.

Chroma (ChromaDB) is an open-source, AI-native search and vector database that enables developers to store, index, and retrieve high-dimensional embeddings for LLM applications. Its core database product supports hybrid retrieval—combining dense vector similarity, sparse (BM25/SPLADE) keyword, full-text, regex, and metadata search—through a simple Python, JavaScript/TypeScript, or Rust SDK. Chroma Cloud, the managed serverless offering GA since August 2025, is built on object storage (S3/GCS) with intelligent caching and tiering, SOC 2 Type II compliance, and a BYOC enterprise option. Complementary products include Chroma Sync (automated data ingestion from GitHub and web), Chroma Agent (self-editing search agent research project), and Package Search MCP for AI agent tool use.

Trieve was a source-available, API-first infrastructure platform for search, RAG, recommendations, and analytics. It abstracted the complexity of combining vector databases, embedding models, rerankers, and LLM inference into a 72-endpoint REST API, with SDKs for TypeScript and Python, self-hosting guides for major cloud providers, and front-end UI components for search and merchandising. Following acquisition by Mintlify in July 2025, Trieve's commercial cloud was sunset and its codebase open-sourced under the MIT license.