Pinecone MCP server

What the model can do

Pinecone vector database: indexes with their type, region, size, model and protection, record counts per namespace, namespaces, backups, restore jobs and bulk imports; when you ask, can configure an index (replicas, pod size, deletion protection, tags), back it up, restore a backup to a new index, delete an index or backup, create or delete a namespace, delete records by id, update a record's metadata, and cancel an import.

Reads

• pinecone_list_indexesPinecone indexes in the project: name, state, serverless or pod (cloud, region, pod type, replicas), dimension, metric, embedding model…
• pinecone_describe_indexOne Pinecone index's configuration and status.
• pinecone_index_statsHow many records a Pinecone index holds, in total and per namespace (largest first), with its dimension and, for pod indexes, how full it is.
• pinecone_list_namespacesNamespaces in a Pinecone index with their record counts and indexed metadata field names.
• pinecone_list_backupsPinecone backups, for one index or the whole project: name, status, source index, region, record and namespace counts, size and when taken.
• pinecone_list_restore_jobsPinecone restore jobs (a backup being restored into a new index): status, progress, backup and target index.
• pinecone_list_importsBulk imports into a Pinecone index: status, progress, records imported, source and any error.

Acts, when you ask

These change something in Pinecone. The model is told to use them only when you ask for exactly that, and each one is granted by name: per team on a shared connection, by the owner on a personal one.

• pinecone_configure_indexChanges a Pinecone index's configuration: replicas or pod type of a pod index (changes what it costs), deletion protection…
• pinecone_create_backupTakes a backup of a Pinecone index; backup storage is billed.
• pinecone_restore_backupCreates a new Pinecone index from a backup; the new index is billed like any other.
• pinecone_delete_indexPermanently deletes a Pinecone index and every record in it; it cannot be undone.
• pinecone_delete_backupPermanently deletes a Pinecone backup; it cannot be undone.
• pinecone_create_namespaceCreates an empty namespace in a Pinecone serverless index.
• pinecone_delete_namespacePermanently deletes a namespace from a Pinecone serverless index together with every record in it; it cannot be undone.
• pinecone_delete_recordsPermanently deletes the records with the ids given (at most 1000) from one namespace of a Pinecone index; it cannot be undone.
• pinecone_update_record_metadataSets metadata fields on one record (by id) in a Pinecone index; the fields given replace those values, other fields and the vector stay.
• pinecone_cancel_importCancels a running bulk import into a Pinecone index; records already imported stay.

Connect it

First connect Pinecone on Dashboard → Connections, then add Memtro to your client once. The Pinecone tools appear alongside the tools of everything else you have connected; nothing is configured per connector.

claude plugin marketplace add memtro-org/claude-memtro
claude plugin install memtro@memtro --config api_key=sk-memtro-...

{
  "mcpServers": {
    "memtro": {
      "url": "https://platform.memtro.com/api/mcp",
      "headers": {
        "Authorization": "Bearer sk-memtro-..."
      }
    }
  }
}

{
  "mcp": {
    "memtro": {
      "type": "remote",
      "url": "https://platform.memtro.com/api/mcp",
      "headers": {
        "Authorization": "Bearer sk-memtro-..."
      }
    }
  }
}

Who may use it

The key identifies the person, and the server offers that person the tools they may use everywhere else in Memtro. A shared Pinecone connection is used by the teams an administrator granted it to, tool by tool, on the Permissions page; a personal connection is used by its owner alone, and its tools that act stay off until the owner allows them. See Teams and permissions and Connectors.

Ask it like this

• › How many records does each index hold?
• › Which namespaces are largest in the support-docs index?
• › When was the products index last backed up?
• › Did last night's bulk import finish, and were there errors?
• › Turn on deletion protection for the production index.

Questions

Is this an official Pinecone MCP server?

No. It is Memtro's MCP server with Memtro's Pinecone connector behind it: the connector reads Pinecone through its API with the credentials you connect on Dashboard → Connections, and Pinecone publishes nothing here. The same endpoint carries every other system you connect and Memtro's memory tools.

What does it cost?

Nothing on top of Memtro. One price per organisation by its number of users, from $50 a month for up to 10 users after a 7-day free trial, with every connector, the MCP server, jobs and memory included. You bring your own model keys.

Can it change anything in Pinecone?

Only when you ask, and only what it has been granted: configure an index: replicas, pod size, deletion protection or tags; back up an index; restore a backup into a new index; delete an index; delete a backup; create a namespace in an index; delete a namespace and all its records; delete records by id; update a record's metadata; cancel a bulk import. Every other tool reads, searches or lists. An administrator grants each acting tool by name per team, and a personal connection acts only after its owner allows it.

Which clients work?

Any MCP client that speaks Streamable HTTP: Claude Code, Cursor, Claude Desktop, OpenCode, Codex, Zed, Windsurf, Cline and the Anthropic and OpenAI SDKs. There is no local process to run; the server is remote and authenticated with your key.

Other MCP servers · Databases and warehouses

• Postgres MCP server
• MySQL and MariaDB MCP server
• SQL Server MCP server
• BigQuery MCP server
• Snowflake MCP server
• AWS Athena MCP server