Pinecone MCP

Pinecone MCP Tools Reference

The Pinecone MCP server exposes the following tools to AI agents and IDEs. For setup and installation instructions, see the MCP server guide (opens in a new tab).

Key Limitation: The Pinecone MCP only supports integrated indexes — indexes created with a built-in Pinecone embedding model. It does not work with standard indexes using external embedding models. For those, use the Pinecone CLI.

list-indexes

List all indexes in the current Pinecone project.

describe-index

Get configuration details for a specific index — cloud, region, dimension, metric, embedding model, field map, and status.

Parameters:

• name (required) — Index name

describe-index-stats

Get statistics for an index including total record count and per-namespace breakdown.

Parameters:

• name (required) — Index name

create-index-for-model

Create a new serverless index with an integrated embedding model. Pinecone handles embedding automatically — no external model needed.

Parameters:

• name (required) — Index name
• cloud (required) — aws, gcp, or azure
• region (required) — Cloud region (e.g. us-east-1)
• embed.model (required) — Embedding model: llama-text-embed-v2, multilingual-e5-large, or pinecone-sparse-english-v0
• embed.fieldMap.text (required) — The record field that contains text to embed (e.g. chunk_text)

upsert-records

Insert or update records in an integrated index. Records are automatically embedded using the index’s configured model.

Parameters:

• name (required) — Index name
• namespace (required) — Namespace to upsert into
• records (required) — Array of records. Each record must have an id or _id field and contain the text field specified in the index’s fieldMap. Do not nest fields under metadata — put them directly on the record.

Example record:

{ "_id": "rec1", "chunk_text": "The Eiffel Tower was built in 1889.", "category": "architecture" }

search-records

Semantic text search against an integrated index. Pass plain text — the MCP embeds the query automatically using the index’s model.

Parameters:

• name (required) — Index name
• namespace (required) — Namespace to search
• query.inputs.text (required) — The text query
• query.topK (required) — Number of results to return
• query.filter (optional) — Metadata filter using MongoDB-style operators ($eq, $ne, $in, $gt, $gte, $lt, $lte)
• rerank.model (optional) — Reranking model: bge-reranker-v2-m3, cohere-rerank-3.5, or pinecone-rerank-v0
• rerank.rankFields (optional) — Fields to rerank on (e.g. ["chunk_text"])
• rerank.topN (optional) — Number of results to return after reranking

cascading-search

Search across multiple indexes simultaneously, then deduplicate and rerank results into a single ranked list.

Parameters:

• indexes (required) — Array of { name, namespace } objects to search across
• query.inputs.text (required) — The text query
• query.topK (required) — Number of results to retrieve per index before reranking
• rerank.model (required) — Reranking model: bge-reranker-v2-m3, cohere-rerank-3.5, or pinecone-rerank-v0
• rerank.rankFields (required) — Fields to rerank on
• rerank.topN (optional) — Final number of results to return after reranking

rerank-documents

Rerank a set of documents or records against a query without performing a vector search first.

Parameters:

• model (required) — bge-reranker-v2-m3, cohere-rerank-3.5, or pinecone-rerank-v0
• query (required) — The query to rerank against
• documents (required) — Array of strings or records to rerank
• options.topN (required) — Number of results to return
• options.rankFields (optional) — If documents are records, the field(s) to rerank on