Python Client

This page provides installation instructions, usage examples, and a reference for the Pinecone Python client.

Getting Started

Installation

Use the following shell command to install the Python client for use with Python versions 3.6+:

pip3 install pinecone-client

Alternatively, you can install Pinecone in a Jupyter notebook:

!pip3 install pinecone-client

We strongly recommend installing Pinecone in a virtual environment. For more information on using Python virtual environments, see:

There is a gRPC flavor of the client available, which comes with more dependencies in return for faster upload speeds. To install it, use the following command:

pip3 install "pinecone-client[grpc]"

For the latest development version:

pip3 install git+https://[email protected]/pinecone-io/pinecone-python-client.git

For a specific development version:

pip3 install git+https://[email protected]/pinecone-io/pinecone-python-client.git
pip3 install git+https://[email protected]/pinecone-io/[email protected]
pip3 install git+https://[email protected]/pinecone-io/[email protected]

Usage

Create index

The following example creates an index without a metadata configuration. By default, Pinecone indexes all metadata.

import pinecone

pinecone.init(api_key="YOUR_API_KEY",
              environment="YOUR_ENVIRONMENT")

pinecone.create_index("example-index", dimension=1024)

The following example creates an index that only indexes the "color" metadata field. Queries against this index cannot filter based on any other metadata field.

metadata_config = {
    "indexed": ["color"]
}

pinecone.create_index("example-index-2", dimension=1024,
                      metadata_config=metadata_config)

List indexes

The following example returns all indexes in your project.

import pinecone

pinecone.init(api_key="YOUR_API_KEY", environment="YOUR_ENVIRONMENT")

active_indexes = pinecone.list_indexes()

Describe index

The following example returns information about the index example-index.

import pinecone

pinecone.init(api_key="YOUR_API_KEY", environment="YOUR_ENVIRONMENT")

index_description = pinecone.describe_index("example-index")

Delete index

The following example deletes example-index.

import pinecone

pinecone.init(api_key="YOUR_API_KEY", environment="YOUR_ENVIRONMENT")

pinecone.delete_index("example-index")

Scale replicas

The following example changes the number of replicas for example-index.

import pinecone

pinecone.init(api_key="YOUR_API_KEY", environment="YOUR_ENVIRONMENT")

new_number_of_replicas = 4
pinecone.configure_index("example-index", replicas=new_number_of_replicas)

Describe index statistics

The following example returns statistics about the index example-index.

import pinecone

pinecone.init(api_key="YOUR_API_KEY", environment="YOUR_ENVIRONMENT")
index = pinecone.Index("example-index")

index_stats_response = index.describe_index_stats()

Upsert vectors

The following example upserts dense vectors to example-index.

import pinecone

pinecone.init(api_key="YOUR_API_KEY", environment="YOUR_ENVIRONMENT")
index = pinecone.Index("example-index")

upsert_response = index.upsert(
    vectors=[
        (
         "vec1",                # Vector ID 
         [0.1, 0.2, 0.3, 0.4],  # Dense vector values
         {"genre": "drama"}     # Vector metadata
        ),
        (
         "vec2", 
         [0.2, 0.3, 0.4, 0.5], 
         {"genre": "action"}
        )
    ],
    namespace="example-namespace"
)

Query an index

The following example queries the index example-index with metadata filtering.

import pinecone

pinecone.init(api_key="YOUR_API_KEY", environment="YOUR_ENVIRONMENT")
index = pinecone.Index("example-index")

query_response = index.query(
    namespace="example-namespace",
    top_k=10,
    include_values=True,
    include_metadata=True,
    vector=[0.1, 0.2, 0.3, 0.4],
    filter={
        "genre": {"$in": ["comedy", "documentary", "drama"]}
    }
)

Delete vectors

The following example deletes vectors by ID.

import pinecone

pinecone.init(api_key="YOUR_API_KEY", environment="YOUR_ENVIRONMENT")
index = pinecone.Index("example-index")

delete_response = index.delete(ids=["vec1", "vec2"], namespace="example-namespace")

Fetch vectors

The following example fetches vectors by ID.

import pinecone

pinecone.init(api_key="YOUR_API_KEY", environment="YOUR_ENVIRONMENT")
index = pinecone.Index("example-index")

fetch_response = index.fetch(ids=["vec1", "vec2"], namespace="example-namespace")

Update vectors

The following example updates vectors by ID.

import pinecone

pinecone.init(api_key="YOUR_API_KEY", environment="YOUR_ENVIRONMENT")
index = pinecone.Index("example-index")

update_response = index.update(
    id="vec1",
    values=[0.1, 0.2, 0.3, 0.4],
    set_metadata={"genre": "drama"},
    namespace="example-namespace"
)

Create collection

The following example creates the collection example-collection from example-index.

import pinecone

pinecone.init(api_key="YOUR_API_KEY",
              environment="YOUR_ENVIRONMENT")

pinecone.create_collection("example-collection", "example-index")

List collections

The following example returns a list of the collections in the current project.

import pinecone

pinecone.init(api_key="YOUR_API_KEY", environment="YOUR_ENVIRONMENT")

active_collections = pinecone.list_collections()

Describe a collection

The following example returns a description of the collection example-collection.

import pinecone

pinecone.init(api_key="YOUR_API_KEY", environment="YOUR_ENVIRONMENT")

collection_description = pinecone.describe_collection("example-collection")

Delete a collection

The following example deletes the collection example-collection.

import pinecone

pinecone.init(api_key="YOUR_API_KEY", environment="YOUR_ENVIRONMENT")

pinecone.delete_collection("example-collection")

Reference

For the REST API or other clients, see the API reference.

init()

pinecone.init(**kwargs)

Initialize Pinecone.

Parameters	Type	Description
`api_key`	str	Your Pinecone API key.
`environment`	str	The cloud environment of your Pinecone project.

Example:

import pinecone
pinecone.init(api_key="YOUR_API_KEY", environment="YOUR_ENVIRONMENT")

configure_index()

pinecone.configure_index(index_name, **kwargs)

Configure an index to change pod type and number of replicas.

Parameters	Type	Description
`index_name`	str	The name of the index
`replicas`	int	(Optional) The number of replicas to configure for this index.
`pod_type`	str	(Optional) The new pod type for the index. One of `s1`, `p1`, or `p2` appended with `.` and one of `x1`, `x2`, `x4`, or `x8`.

Example:

import pinecone

pinecone.init(api_key='YOUR_API_KEY', environment='YOUR_ENVIRONMENT')

new_number_of_replicas = 4
pinecone.configure_index('example-index', replicas=new_number_of_replicas)

create_collection()

pinecone.create_collection(**kwargs)

Create a collection from an index.

Parameters	Type	Description
`name`	str	The name of the collection to be created.
`source`	str	The name of the source index to be used as the source for the collection.

Example:

import pinecone

pinecone.init(api_key='YOUR_API_KEY', environment='YOUR_ENVIRONMENT')

pinecone.create_collection('example-collection', 'example-index')

create_index()

pinecone.create_index(**kwargs)

Create an index.

Parameters	Type	Description
`name`	str	The name of the index to be created. The maximum length is 45 characters.
`dimension`	integer	The dimensions of the vectors to be inserted in the index.
`metric`	str	(Optional) The distance metric to be used for similarity search: 'euclidean', 'cosine', or 'dotproduct'.
`pods`	int	(Optional) The number of pods for the index to use, including replicas.
`replicas`	int	(Optional) The number of replicas.
`pod_type`	str	(Optional) The new pod type for the index. One of `s1`, `p1`, or `p2` appended with `.` and one of `x1`, `x2`, `x4`, or `x8`.
`metadata_config`	object	(Optional) Configuration for the behavior of Pinecone's internal metadata index. By default, all metadata is indexed; when metadata_config is present, only specified metadata fields are indexed. To specify metadata fields to index, provide a JSON object of the following form: `{"indexed": ["example_metadata_field"]}`
`source_collection`	str	(Optional) The name of the collection to create an index from.

Example:

import pinecone

pinecone.init(api_key='YOUR_API_KEY', environment='YOUR_ENVIRONMENT')

## The following example creates an index without a metadata
## configuration. By default, Pinecone indexes all metadata.

pinecone.create_index('example-index', dimension=1024)

## The following example creates an index that only indexes
## the 'color' metadata field. Queries against this index
## cannot filter based on any other metadata field.

metadata_config = {
    'indexed': ['color']
}

pinecone.create_index('example-index-2', dimension=1024,
                      metadata_config=metadata_config)

delete_collection()

pinecone.delete_collection('example-collection')

Delete an existing collection.

Parameters	Type	Description
`collectionName`	str	The name of the collection to delete.

Example:

import pinecone

pinecone.init(api_key='YOUR_API_KEY', environment='YOUR_ENVIRONMENT')
pinecone.delete_collection('example-collection')

delete_index()

pinecone.delete_index(indexName)

Delete an existing index.

Parameters	Type	Description
`index_name`	str	The name of the index.

Example:

import pinecone

pinecone.init(api_key='YOUR_API_KEY', environment='YOUR_ENVIRONMENT')

pinecone.delete_index('example-index')

describe_collection()

pinecone.describe_collection(collectionName)

Get a description of a collection.

Parameters	Type	Description
`collection_name`	str	The name of the collection.

Example:

import pinecone

pinecone.init(api_key='YOUR_API_KEY', environment='YOUR_ENVIRONMENT')

collection_description = pinecone.describe_collection('example-collection')

Returns:

collectionMeta : object Configuration information and deployment status of the collection.
- name : string The name of the collection.
- size: integer The size of the collection in bytes.
- status: string The status of the collection.

describe_index()

pinecone.describe_index(indexName)

Get a description of an index.

Parameters	Type	Description
`index_name`	str	The name of the index.

Returns:

database : object
name : string The name of the index.
dimension : integer The dimensions of the vectors to be inserted in the index.
metric : string The distance metric used for similarity search: 'euclidean', 'cosine', or 'dotproduct'.
pods : integer The number of pods the index uses, including replicas.
replicas : integer The number of replicas.
pod_type : string The pod type for the index. One of s1, p1, or p2 appended with . and one of x1, x2, x4, or x8.
metadata_config: object Configuration for the behavior of Pinecone's internal metadata index. By default, all metadata is indexed; when metadata_config is present, only specified metadata fields are indexed. To specify metadata fields to index, provide a JSON object of the following form: {"indexed": ["example_metadata_field"]}
status : object
ready : boolean Whether the index is ready to serve queries.
state : string One of Initializing, ScalingUp, ScalingDown, Terminating, or Ready.

Example:

import pinecone

pinecone.init(api_key='YOUR_API_KEY', environment='YOUR_ENVIRONMENT')

index_description = pinecone.describe_index('example-index')

list_collections()

pinecone.list_collections()

Return a list of the collections in your project.

Returns:

array of strings The names of the collections in your project.

Example:

import pinecone

pinecone.init(api_key='YOUR_API_KEY', environment='us-east1-gcp')

active_collections = pinecone.list_collections()

list_indexes()

pinecone.list_indexes()

Return a list of your Pinecone indexes.

Returns:

array of strings The names of the indexes in your project.

Example:

import pinecone

pinecone.init(api_key='YOUR_API_KEY', environment='YOUR_ENVIRONMENT')

active_indexes = pinecone.list_indexes()

Index()

pinecone.Index(indexName)

Construct an Index object.

Parameters	Type	Description
`indexName`	str	The name of the index.

Example:

index = pinecone.Index("example-index")

Index.delete()

Index.delete(**kwargs)

Delete items by their ID from a single namespace.

Parameters	Type	Description
`ids`	array	(Optional) `array` of `strings` vectors to delete.
`delete_all`	boolean	(Optional) Indicates that all vectors in the index namespace should be deleted.
`namespace`	str	(Optional) The namespace to delete vectors from, if applicable.
`filter`	object	(Optional) If specified, the metadata filter here will be used to select the vectors to delete. This is mutually exclusive with specifying ids to delete in the ids param or using `delete_all=True`. See https://www.pinecone.io/docs/metadata-filtering/.

Example:

import pinecone

pinecone.init(api_key='YOUR_API_KEY', environment='YOUR_ENVIRONMENT')
index = pinecone.Index('example-index')

delete_response = index.delete(ids=['vec1', 'vec2'], namespace='example-namespace')

Index.describe_index_stats()

Index.describe_index_stats()

Returns statistics about the index's contents, including the vector count per namespace and the number of dimensions.

Returns:

namespaces : object A mapping for each namespace in the index from the namespace name to a
summary of its contents. If a metadata filter expression is present, the summary will reflect only vectors matching that expression.
dimension : int64 The dimension of the indexed vectors.
indexFullness : float The fullness of the index, regardless of whether a metadata filter expression was passed. The granularity of this metric is 10%.
totalVectorCount : int64 The total number of vectors in the index.

Example:

import pinecone

pinecone.init(api_key='YOUR_API_KEY', environment='YOUR_ENVIRONMENT')
index = pinecone.Index('example-index')

index_stats_response = index.describe_index_stats()

Index.fetch()

Index.fetch(ids, **kwargs)

The Fetch operation looks up and returns vectors, by ID, from a single namespace. The returned vectors include the vector data and metadata.

Parameters	Type	Description
`ids`	[str]	The vector IDs to fetch. Does not accept values containing spaces.
`namespace`	str	(Optional) The namespace containing the vectors.

Returns:

vectors : object Contains the vectors.
namespace : string The namespace of the vectors.

Example:

import pinecone

pinecone.init(api_key='YOUR_API_KEY', environment='YOUR_ENVIRONMENT')
index = pinecone.Index('example-index')

fetch_response = index.fetch(ids=['vec1', 'vec2'], namespace='example-namespace')

Index.query()

Index.query(**kwargs)

Search a namespace using a query vector. Retrieves the ids of the most similar items in a namespace, along with their similarity scores.

Parameters	Type	Description
`namespace`	str	(Optional) The namespace to query.
`top_k`	int64	The number of results to return for each query.
`filter`	object	(Optional) The filter to apply. You can use vector metadata to limit your search. See https://www.pinecone.io/docs/metadata-filtering/.
`include_values`	boolean	(Optional) Indicates whether vector values are included in the response. Defaults to `false`.
`include_metadata`	boolean	(Optional) Indicates whether metadata is included in the response as well as the ids. Defaults to `false`.
`vector`	[floats]	(Optional) The query vector. This should be the same length as the dimension of the index being queried. Each `query()` request can contain only one of the parameters `id` or `vector`.
`sparse_vector`	dictionary	(Optional) The sparse query vector. This must contain an array of integers named `indices` and an array of floats named `values`. These two arrays must be the same length.
`id`	`string`	(Optional) The unique ID of the vector to be used as a query vector. Each `query()` request can contain only one of the parameters `vector` or `id`.

Example:

import pinecone

 pinecone.init(api_key='YOUR_API_KEY', environment='YOUR_ENVIRONMENT')
 index = pinecone.Index('example-index')

 query_response = index.query(
    namespace='example-namespace',
    top_k=10,
    include_values=True,
    include_metadata=True,
    vector=[0.1, 0.2, 0.3, 0.4],
    filter={
        'genre': {'$in': ['comedy', 'documentary', 'drama']}
    }
)

The following example queries the index example-index with a sparse-dense vector.

import pinecone

pinecone.init(api_key="YOUR_API_KEY", environment="YOUR_ENVIRONMENT")
index = pinecone.Index("example-index")

query_response = index.query(
    namespace="example-namespace",
    top_k=10,
    include_values=True,
    include_metadata=True,
    vector=[0.1, 0.2, 0.3, 0.4],
    sparse_vector={
        'indices': [10, 45, 16],
        'values':  [0.5, 0.5, 0.2]
    },
    filter={
        "genre": {"$in": ["comedy", "documentary", "drama"]}
    }
)

Index.update()

Index.update(**kwargs)

Updates vectors in a namespace. If a value is included, it will overwrite the previous value.
If set_metadata is included, the values of the fields specified in it will be added or overwrite the previous value.

Parameters	Type	Description
`id`	str	The vector's unique ID.
`values`	[float]	(Optional) Vector data.
`set_metadata`	object	(Optional) Metadata to set for the vector.
`namespace`	str	(Optional) The namespace containing the vector.

Example:

import pinecone

pinecone.init(api_key='YOUR_API_KEY', environment='YOUR_ENVIRONMENT')
index = pinecone.Index('example-index')

update_response = index.update(
    id='vec1',
    values=[0.1, 0.2, 0.3, 0.4],
    set_metadata={'genre': 'drama'},
    namespace='example-namespace'
)

Index.upsert()

Index.upsert(**kwargs)

Writes vectors into a namespace. If a new value is upserted for an existing vector ID, it will overwrite the previous value.

Parameters	Type	Description
`vectors`	[object]	An array containing the vectors to upsert. Recommended batch limit is 100 vectors. `id` (str) - The vector's unique id. `values` ([float]) - The vector data. `metadata` (object) - (Optional) Metadata for the vector. `sparse_vector` (object) - (Optional) A dictionary containing the index and values arrays containing the sparse vector values.
`namespace`	str	(Optional) The namespace name to upsert vectors.

Returns:

upsertedCount : int64 The number of vectors upserted.

Example:

import pinecone

pinecone.init(api_key='YOUR_API_KEY', environment='YOUR_ENVIRONMENT')
index = pinecone.Index('example-index')

upsert_response = index.upsert(
   vectors=[
       {'id': "vec1", "values":[0.1, 0.2, 0.3, 0.4], "metadata": {'genre': 'drama'}},
       {'id': "vec2", "values":[0.2, 0.3, 0.4, 0.5], "metadata": {'genre': 'action'}},
   ],
   namespace='example-namespace'
)

The following example upserts vectors with sparse and dense values to example-index.

import pinecone

pinecone.init(api_key="YOUR_API_KEY", environment="YOUR_ENVIRONMENT")
index = pinecone.Index("example-index")

upsert_response = index.upsert(
    vectors=[
        {'id': 'vec1',
         'values': [0.1, 0.2, 0.3, 0.4],
         'metadata': {'genre': 'drama'},
         'sparse_values': {
             'indices': [10, 45, 16],
             'values': [0.5, 0.5, 0.2]
         }},
        {'id': 'vec2',
         'values': [0.2, 0.3, 0.4, 0.5],
         'metadata': {'genre': 'action'},
         'sparse_values': {
             'indices': [15, 40, 11],
             'values': [0.4, 0.5, 0.2]
         }}
    ],
    namespace='example-namespace'
)