hermes

Connect Hermes to Pinecone

Give an AI agent your Pinecone as a tool and it can build and query a vector store on its own instead of just reasoning about one. It can create an index, generate embeddings, upload and index documents, run a similarity query, and rerank the matches to surface the best passages, calling each operation as a tool while it works a task. That turns an agent into something that maintains its own retrieval layer, writing new records and reading them back mid-task.

Loading preview…
Opens a workflow with 22 tools ready to use.

Tools Hermes gets

Pick the operations you want and each becomes a tool the agent can call directly while it works.

Pinecone

22 tools
Query / Search

Find the most similar vectors to a query vector (similarity search).

Upsert Vectors

Insert or overwrite vectors in an index.

Fetch Vectors

Fetch vectors by their IDs.

Delete Vectors

Delete vectors by ID, by metadata filter, or clear a namespace.

Update Vector

Partially update values or metadata on a single existing vector.

List Vectors

List vector IDs in a namespace by prefix (serverless indexes only).

Describe Index Stats

Get per-namespace vector counts, total dimension, and index fullness.

List Indexes

List all indexes in the project.

Describe Index

Get full details for a single index: status, host, metric, dimension, spec.

Create Index

Create a new serverless index.

Configure Index

Change deletion protection or pod replica/type on an existing index.

Delete Index

Delete an index and all its vectors permanently.

Create Index for Model

Create a text-native index that embeds content server-side (no dimension needed).

Upsert Records (Text-Native)

Upsert text records into a text-native index (server-side embedding, no vectors needed).

Search Records (Text-Native)

Search a text-native index with a text query (server-side embedding + optional rerank).

Generate Embeddings

Generate dense embeddings using Pinecone's hosted models.

Rerank Documents

Rerank a set of documents against a query using Pinecone's reranking model.

Start Import

Start an async bulk import of Parquet vectors from S3, GCS, or Azure Blob.

List Imports

List all bulk import operations for an index.

Describe Import

Get the status and details of a specific bulk import operation.

Cancel Import

Cancel a pending or in-progress bulk import operation.

Upload & Index Document

Upload a document file, chunk + embed it with text-embedding-3-small, and upsert.

What you can build

When a new document lands from an upstream node, generate embeddings and upload it into a Pinecone index so it becomes searchable right away.
Answer a user's question by running a Query / Search against an index, then rerank the matches to feed only the most relevant passages into an AI step.
On a schedule, describe index stats to check vector counts and alert the team when an index drifts from its expected size.
Keep records current by fetching a vector, updating its values or metadata, and writing the change back without rebuilding the whole index.
Load a large dataset by starting a bulk import, then list and describe imports to track progress until it finishes.
Prune stale content by deleting specific vectors from an index whenever the source records are removed elsewhere in the workflow.

About Hermes

Hermes is an open agent built on Nous Research’s Hermes models. NoClick runs it hosted and connects your apps to it as tools, so it can do real work across your systems. Wire an integration into the agent and Hermes can use that app’s operations directly while it reasons through a task. It is a strong pick when you want an open-model agent with genuine tool access.

Built on open Hermes models
Capable general reasoning
Flexible, open-model foundation
Hosted with apps wired in as tools

Frequently asked questions

Build a custom Hermes agentPinecone integration

Open your Hermes agent in one click

NoClick runs Hermes for you with Pinecone wired in as tools. Connect your account and run.