Pinecone is a Vector Databases company founded in 2019 and based in New York, United States. It has raised $138M in total funding, most recently a Series B in 2023 at a $750M valuation.
| Date | Stage | Amount | Valuation | Lead investors |
|---|---|---|---|---|
| Apr 27, 2023 | Series B | $100M | $750M | Andreessen Horowitz |
| Mar 1, 2022 | Series A | $28M | — |
| Menlo Ventures |
| Jan 1, 2021 | Seed | $10M | — | Wing Venture Capital |
A fully managed, serverless vector database that automatically scales read and write capacity based on demand without requiring cluster provisioning. Designed for production AI applications including RAG pipelines and recommendation systems. Technically notable for separating storage and compute to minimize costs at rest while maintaining sub-millisecond query latency at query time across billions of vectors.
A hosted service enabling developers to build production-grade retrieval-augmented chat and agentic applications without managing embedding pipelines. Accepts documents, generates embeddings, stores them in Pinecone indexes, and returns grounded responses from an LLM. Targets teams that want a managed end-to-end RAG stack rather than assembling components individually.
A managed embedding and reranking inference service integrated directly into the Pinecone pipeline. Allows users to generate and update vector embeddings from text without managing external model hosting. Supports leading embedding models and reranking models to improve retrieval precision, eliminating the operational overhead of running inference infrastructure separately from the vector store.
13 patents on file, but none with both an extractable figure and an abstract on Google Patents yet.