Pinecone

Pinecone is a vector database — the storage layer that powers semantic search and retrieval for AI apps, holding your data as embeddings so an assistant can find records by meaning rather than exact keywords. The official connector, published by Pinecone, is the Pinecone Developer MCP Server: it lets coding assistants and other AI tools connect to your Pinecone projects and documentation. Once connected, the AI can search the Pinecone docs to answer questions accurately, help you configure indexes based on your application's needs, generate code informed by your index configuration and data, and upsert and search records so you can test queries and evaluate results right in your dev environment.

Ask "list all my Pinecone indexes and describe their configurations" and it reads your project and answers. Say "create a new index called 'my-docs' using the multilingual-e5-large model" and it sets it up. It talks to Pinecone with your API key, so it can do what you can do from your own account. Note: without an API key the AI can still search the documentation — it just can't manage or query your indexes.

What you can ask

• "Search the Pinecone docs for information about metadata filtering."
• "List all my Pinecone indexes and describe their configurations."
• "Create a new index called 'my-docs' using the multilingual-e5-large model."
• "Upsert these documents into my index: [paste your documents]."
• "Search my index for records related to 'authentication best practices'."
• "What namespaces exist in my index, and how many records are in each?"

What token you need

You need a Pinecone API key — the key that lets software act on your Pinecone project.

1. Sign in to the Pinecone console and generate an API key.
2. Copy it.
3. Paste it when the Lab asks for PINECONE_API_KEY. It's stored encrypted in your Keychain.

The API key can create, change, and query real indexes and data, so treat it like a password. A word of honesty: without an API key the server still works for documentation search, but it won't be able to manage or query your indexes. The server also only supports indexes with integrated inference — see the limitation below.