Manage a dedicated read nodes index

Use a hosted embedding model

To upsert and search with text instead of vectors, an index must use a hosted embedding model. On 2026-07, an index schema is immutable, so you set this when you create the index, not after. To create a dedicated read nodes index with a hosted embedding model, call Create an index with integrated embedding and include the read_capacity object to set the node type, shards, and replicas.

Monitor index fullness

To check index fullness, call Get index stats.

Example

# To get the unique host for an index,
# see https://docs.pinecone.io/guides/manage-data/target-an-index
PINECONE_API_KEY="YOUR_API_KEY"
INDEX_HOST="YOUR_INDEX_HOST"

curl -X GET "https://$INDEX_HOST/describe_index_stats" \
     -H "Api-Key: $PINECONE_API_KEY" \
     -H "X-Pinecone-Api-Version: 2026-07"

{
  "namespaces": {
    "__default__": {
      "vectorCount": 705000
    }
  },
  "indexFullness": 0.01,
  "totalVectorCount": 705000,
  "dimension": 1536,
  "metric": "cosine",
  "vectorType": "dense",
  "memoryFullness": 0.01,
  "storageFullness": 0.01
}

In the response, indexFullness describes how full the index is, on a scale of 0 to 1. It’s set to the greater of memoryFullness and storageFullness. Pinecone also emits these values. Use them to track fullness over time in Prometheus or Datadog, and to alert before your index reaches capacity.

Change node types

You can change node types in either direction (b1 → t1 or t1 → b1). This operation doesn’t require downtime, but can take up to 30 minutes to complete.

The most predictable way to increase throughput is by increasing replicas.

t1 nodes cache more data in memory than b1 nodes. Because of this, switching from b1 to t1 may require more shards.If your new configuration doesn’t have enough shards, the configuration change will fail with an error telling you how many shards are required. Update the request and retry.In the meantime, your index will continue to function normally in its original configuration.

To change node types, call Configure an index. In the request body, set the following fields:

Field	Value	Notes
`read_capacity.mode`	`Dedicated`
`read_capacity.dedicated.node_type`	`b1` or `t1`	See node types

Example

This example changes the node type from b1 to t1:

PINECONE_API_KEY="YOUR_API_KEY"
INDEX_NAME="YOUR_INDEX_NAME"

curl -X PATCH "https://api.pinecone.io/indexes/$INDEX_NAME" \
     -H "Accept: application/json" \
     -H "Content-Type: application/json" \
     -H "Api-Key: $PINECONE_API_KEY" \
     -H "X-Pinecone-Api-Version: 2026-07" \
     -d '{
           "read_capacity": {
             "mode": "Dedicated",
             "dedicated": {
               "node_type": "t1"
             }
           }
         }'

{
  "name": "example-dedicated-index",
  "status": {
    "ready": true,
    "state": "Ready"
  },
  "host": "example-dedicated-index-1c6ab6aa.svc.aped-4627-b74a.pinecone.io",
  "deployment": {
    "deployment_type": "managed",
    "region": "us-east-1",
    "cloud": "aws",
    "environment": "aped-4627-b74a"
  },
  "read_capacity": {
    "mode": "Dedicated",
    "dedicated": {
      "node_type": "t1",
      "scaling": "Manual",
      "manual": {
        "shards": 1,
        "replicas": 1
      }
    },
    "status": {
      "state": "Scaling",
      "current_shards": 1,
      "current_replicas": 1
    }
  },
  "schema": {
    "fields": {
      "_values": { "type": "dense_vector", "dimension": 1024, "metric": "cosine" }
    }
  },
  "deletion_protection": "disabled",
  "tags": null
}

Configuration change limits:

You can make one configuration change every ten minutes, but you can batch multiple changes (node type, shards, and replicas) in a single request.
A new configuration change can only be initiated after the previous configuration change has completed.
Each configuration change can take up to 30 minutes to complete.
Read and write operations continue normally during configuration changes.

Pause an index

To pause an index, set the number of replicas to 0. This operation can take up to 30 minutes to complete.

While an index is paused, you can’t write to it or read from it. For a paused index, you’re billed for storage, but not for node costs, reads, or writes.

Check the status of a configuration change

After making a configuration change to a dedicated read nodes index (changing shards, replicas, or node type), check the status of the change by calling Describe an index.

Example

PINECONE_API_KEY="YOUR_API_KEY"
INDEX_NAME="YOUR_INDEX_NAME"

curl -X GET "https://api.pinecone.io/indexes/$INDEX_NAME" \
     -H "Api-Key: $PINECONE_API_KEY" \
     -H "X-Pinecone-Api-Version: 2026-07"

{
  "name": "example-dedicated-index",
  "status": {
    "ready": true,
    "state": "Ready"
  },
  "host": "example-dedicated-index-1c6ab6aa.svc.aped-4627-b74a.pinecone.io",
  "deployment": {
    "deployment_type": "managed",
    "region": "us-east-1",
    "cloud": "aws",
    "environment": "aped-4627-b74a"
  },
  "read_capacity": {
    "mode": "Dedicated",
    "dedicated": {
      "node_type": "b1",
      "scaling": "Manual",
      "manual": {
        "shards": 1,
        "replicas": 2 // <---- desired state
      }
    },
    "status": {
      "state": "Scaling",
      "current_shards": 1,
      "current_replicas": 1 // <---- current state
    }
  },
  "schema": {
    "fields": {
      "_values": { "type": "dense_vector", "dimension": 1536, "metric": "cosine" }
    }
  },
  "deletion_protection": "enabled",
  "tags": {
    "tag0": "value0"
  }
}

The response includes two status fields:

Field	Description
`status.state`	Overall index status (for example, `Initializing`, `Ready`, `Terminating`)
`read_capacity.status.state`	Read capacity status (`Migrating`, `Scaling`, `Ready`, `Error`)

When changing node types, shards, or replicas, monitor the read capacity status (read_capacity.status.state). Possible values:

State	Description
`Ready`	The change is complete and the index is ready to serve queries at full capacity.
`Scaling`	A change to the number of shards or replicas is in progress.
`Migrating`	A change to the node type or read capacity is in progress.
`Error`	The operation failed. For migrations to dedicated, this typically means you didn’t allocate enough shards for your index size. Check `error_message` for details, and retry with more shards.

During changes to shards, replicas, and node type, the index-level status (status.state) remains Ready. This is because the index can handle reads and writes while its dedicated read capacity scales.

Convert to on-demand

To convert a dedicated read nodes index back to on-demand, call Configure an index and set read_capacity.mode to OnDemand. This converts the index in place, keeping the same index name and host.

PINECONE_API_KEY="YOUR_API_KEY"
INDEX_NAME="YOUR_INDEX_NAME"

curl -X PATCH "https://api.pinecone.io/indexes/$INDEX_NAME" \
     -H "Accept: application/json" \
     -H "Content-Type: application/json" \
     -H "Api-Key: $PINECONE_API_KEY" \
     -H "X-Pinecone-Api-Version: 2026-07" \
     -d '{
           "read_capacity": {
             "mode": "OnDemand"
           }
         }'

{
  "name": "example-index",
  "status": {
    "ready": true,
    "state": "Ready"
  },
  "host": "example-index-1c6ab6aa.svc.aped-4627-b74a.pinecone.io",
  "deployment": {
    "deployment_type": "managed",
    "region": "us-east-1",
    "cloud": "aws",
    "environment": "aped-4627-b74a"
  },
  "read_capacity": {
    "mode": "OnDemand",
    "status": {
      "state": "Ready",
      "current_shards": null,
      "current_replicas": null
    }
  },
  "schema": {
    "fields": {
      "_values": { "type": "dense_vector", "dimension": 1024, "metric": "cosine" }
    }
  },
  "deletion_protection": "disabled",
  "tags": null
}

If you’d rather return to on-demand on a fresh index, use the backup and restore path instead:

Create a backup of your dedicated read nodes index.
Create a new index from the backup, without specifying dedicated read node configuration.
Verify the new on-demand index and update your application to use it.
Delete the old dedicated read nodes index.

If you have concerns or need assistance, contact support.

Get started

Core concepts

Index data

Search

Optimize

Manage data

Manage cost

Move to production

Admin

Operations

Using pods

Manage a dedicated read nodes index

Use a hosted embedding model

Monitor index fullness

Example

Change node types

Example

Pause an index

Check the status of a configuration change

Example

Convert to on-demand

​Use a hosted embedding model

​Monitor index fullness

​Example

​Change node types

​Example

​Pause an index

​Check the status of a configuration change

​Example

​Convert to on-demand

Use a hosted embedding model

Monitor index fullness

Example

Change node types

Example

Pause an index

Check the status of a configuration change

Example

Convert to on-demand