Weaviate

Weaviate is an AI-native vector database with 16k+ GitHub stars — a Pinecone alternative for semantic search, RAG, and building AI agents with hybrid search capabilities.

Key features

Vector search

• High-performance HNSW indexing
• Hybrid search combining vectors and keywords
• Multi-tenancy with tenant isolation
• Metadata filtering with rich query language
• Multi-vector support per object

AI integration

• Built-in vectorization modules (OpenAI, Cohere, HuggingFace)
• Automatic vector generation and embedding
• RAG pipeline support
• Generative search with LLM integration
• Multi-modal search (text, images)

Developer experience

• GraphQL API for flexible queries
• Python, JavaScript, Go, and Java clients
• Integration with LangChain, LlamaIndex, and Haystack
• Schema management and migrations
• Real-time data updates

Production features

• Horizontal scaling with sharding
• Replication and high availability
• Backup and disaster recovery
• Monitoring and alerting
• Role-based access control

At a glance

License	BSD-3-Clause
Stack	Go, Python, JavaScript
Self-hosted	Yes — Weaviate OSS
Cloud	Weaviate Cloud (managed)
API	GraphQL, REST

Self-hosting

docker run -p 8080:8080 semitechnologies/weaviate

Weaviate can be self-hosted with Docker or Kubernetes. The cloud version provides fully managed infrastructure with automatic scaling.