Quickstart
Pick a search path and run your first search in minutes
Key terms
Organizations, projects, indexes, namespaces, documents, and records
Data modeling
Design the fields your index needs for the searches you’ll run
Search overview
Compare search types and choose the right approach for each query
IDEs & CLIs
Use Pinecone with Claude Code, Codex, Gemini CLI, Cursor, and other agentic tools
MCP server
Connect any MCP-compatible agent to Pinecone for search and index management
What you can do
With Pinecone Database, you can:- Serve full-text search, semantic search, sparse-vector search, and hybrid search from one index. To choose an approach, see Search overview.
- Upsert and query with text instead of vectors. With integrated embedding, Pinecone generates the vectors from your text server-side.
- Narrow results with metadata filters, then rerank them for relevance.
- Keep each tenant’s data separate within one index by using namespaces.
- Scale reads for sustained, high query volumes with dedicated read nodes.
- Run in production by backing up your indexes and deploying in your own cloud account with Bring Your Own Cloud (BYOC).
Resources
API reference
Details about the Pinecone APIs, SDKs, and architecture
Examples
Notebooks and sample apps with common AI patterns
Models
Embedding and reranking models hosted by Pinecone
Integrations
Third-party integrations for LangChain, LlamaIndex, and more
Troubleshooting
Common errors, account help, and how to contact support
Changelog
What’s new in Pinecone
Other Pinecone products
Pinecone Assistant
A managed service for RAG chat and agent applications grounded in your data
Pinecone Nexus
The knowledge engine for agents, serving grounded, cited answers in one call