Release notes

This document contains details about Pinecone releases. For information about using specific features, see our API reference.

March 2024

Features

Improvements

  • Some pod-based indexes on GCP may see a noticeable reduction in index fullness due to storage optimizations.

Console

  • When viewing the list of indexes in a project, you can now search indexes by index name; sort indexes alphabetically, by how recently they were viewed or created, or by status; and filter indexes by index type (serverless, pod-based, or starter).

SDKs

February 2024

Features

  • It is now possible to convert a pod-based starter index to a serverless index. For organizations on the Starter plan, this requires upgrading to Standard or Enterprise; however, upgrading comes with $100 in serverless credits, which will cover the cost of a converted index for some time.

SDKs

Documentation

January 2024

Features

The new Pinecone API gives you the same great vector database but with a drastically improved developer experience. The most significant improvements include:

  • Serverless indexes: With serverless indexes, you don't configure or manage compute and storage resources. You just load your data and your indexes scale automatically based on usage. Likewise, you don't pay for dedicated resources that may sometimes lay idle. Instead, the pricing model for serverless indexes is consumption-based: You pay only for the amount of data stored and operations performed, with no minimums.

    ⚠️

    Warning

    Serverless indexes are in public preview and are available only on AWS in the us-west-2 and us-east-1 regions. Check the current limitations and test thoroughly before using serverless indexes in production.

  • Multi-region projects: Instead of choosing a cloud region for an entire project, you now choose a region for each index in a project. This makes it possible to consolidate related indexes in the same project, even when they are hosted in different regions.

  • Global URL for control plane operations: Control plane operations like create_index, describe_index, and list_indexes now use a single global URL, https://api.pinecone.io, regardless of the cloud environment where an index is hosted. This simplifies the experience compared to the legacy API, where each environment has a unique URL.

SDKs

  • The latest versions of Pinecone's Python client (v3.0.0) and Node.js client (v2.0.0) support the new API mentioned above. To use the new API, existing users must upgrade to the new client versions and adapt some code. For guidance, see the Python client v3 migration guide and Node.js client v2 migration guide.

  • The latest version of the Canopy SDK (v0.6.0) adds support for the new API mentioned above as well as namespaces, LLMs that do not have function calling functionality for query generation, and more. See the release notes in Github for more details.

Documentation

  • The Pinecone documentation is now versioned. The default "latest" version reflects the new Pinecone API mentioned above. The "legacy" version reflects the previous API, which requires regional URLs for control plane operations and does not support serverless indexes.

December 2023

Features

  • The free Starter plan now supports up to 100 namespaces. Namespaces let you partition vectors within an index to speed up queries or comply with multi-tenancy requirements.

November 2023

Features

  • The new Pinecone AWS Reference Architecture is an open-source, distributed system that performs vector-database-enabled semantic search over Postgres records. You can use it as a learning resource or as a starting point for high-scale use cases.

SDKs

  • Canopy is a new open-source Retrieval Augmented Generation (RAG) framework and context engine built on top of Pinecone. It enables you to start chatting with your documents or text data with a few simple commands.

    The latest version of the Canopy SDK (v0.2.0) adds support for OpenAI SDK v1.2.3. See the release notes in Github for more details.

Billing

  • Pinecone is now registered to collect Value Added Tax (VAT) or Goods and Services Tax (GST) for accounts based in various global regions. If applicable, add your VAT or GST number to your account under Settings > Billing.

October 2023

Features

Regions

SDKs

  • The latest version of our Node SDK is v1.1.2. See the release notes in Github for more details.

Console

  • The Index Browser is now available in the console. This allows you to preview, query, and filter by metadata directly from the console. The Index Browser can be found within the index detail page.
  • We’re improved the design of our metrics page to include new charts for record and error count plus additional latencies (p90, p99) to help triage and understand issues.

Integrations

  • Knowledge Base for Amazon Bedrock is now available in Private Preview. Integrate your enterprise data via retrieval augmented generation (RAG) when building search and GenAI applications. Learn more.

  • Pinecone Sink Connector for Confluent is now available in Public Preview. Gain access to data streams from across your business to build a real-time knowledge base for your AI applications. Learn more.

Billing

Privacy

  • Pinecone is now HIPAA compliant across all of our cloud providers (AWS, Azure, and GCP).

September 11, 2023

Pinecone Azure support via the eastus-azure region is now generally available (GA).

August 14, 2023

Pinecone now supports deploying projects to Azure using the new eastus-azure region. This is a public preview environment, so test thoroughly before deploying to production.

June 21, 2023

The new gcp-starter region is now in public preview. This region has distinct limitations from other Starter Plan regions. gcp-starter is the default region for some new users.

April 26, 2023

Indexes in the starter plan now support approximately 100,000 1536-dimensional embeddings with metadata. Capacity is proportional for other dimensionalities.

April 3, 2023

Pinecone now supports new US and EU cloud regions.

March 21, 2023

Pinecone now supports SSO for Enterprise dedicated customers. Contact us at [email protected] to set up your integration.

March 1, 2023

Pinecone now supports 40kb of metadata per vector.

February 22, 2023

Sparse-dense embeddings are now in Public Preview.

Pinecone now supports vectors with sparse and dense values. To use sparse-dense embeddings in Python, upgrade to Python client version 2.2.0.

Pinecone Python client version 2.2.0 is available

Python client version 2.2.0 with support for sparse-dense embeddings is now available on GitHub and PYPI.

February 15, 2023

New Node.js client is now available in public preview

You can now try out our new Node.js client for Pinecone.

February 14, 2023

New usage reports in the Pinecone console

You can now monitor your current and projected Pinecone usage with the Usage dashboard.

January 31, 2023

Pinecone is now available in AWS Marketplace

You can now sign up for Pinecone billing through Amazon Web Services Marketplace.

January 3, 2023

Pinecone Python client version 2.1.0 is now available on GitHub.

The latest release of the Python client makes the following changes:

  • Fixes "Connection Reset by peer" error after long idle periods
  • Adds typing and explicit names for arguments in all client operations
  • Adds docstrings to all client operations
  • Adds Support for batch upserts by passing batch_size to the upsert method
  • Improves gRPC query results parsing performance

December 22, 2022

Pinecone is now available in GCP Marketplace

You can now sign up for Pinecone billing through Google Cloud Platform Marketplace.

December 6, 2022

Organizations are generally available

Pinecone now features organizations, which allow one or more users to control billing and project settings across multiple projects owned by the same organization.

p2 pod type is generally available

The p2 pod type is now generally available and ready for production workloads. p2 pods are now available in the Starter plan and support the dotproduct distance metric.

Performance improvements

October 31, 2022

Hybrid search (Early access)

Pinecone now supports keyword-aware semantic search with the new hybrid search indexes and endpoints. Hybrid search enables improved relevance for semantic search results by combining them with keyword search.

This is an early access feature and is available only by signing up.

October 17, 2022

Status page

The new Pinecone Status Page displays information about the status of the Pinecone service, including the status of individual cloud regions and a log of recent incidents.

September 16, 2022

Public collections

You can now create indexes from public collections, which are collections containing public data from real-world data sources. Currently, public collections include the Glue - SSTB collection, the TREC Question classification collection, and the SQuAD collection.

August 16, 2022

Collections (Public Preview)("Beta")

You can now make static copies of your index using collections. After you create a collection from an index, you can create a new index from that collection. The new index can use any pod type and any number of pods. Collections only consume storage.

This is a public preview feature and is not appropriate for production workloads.

Vertical scaling

You can now change the size of the pods for a live index to accommodate more vectors or queries without interrupting reads or writes. The p1 and s1 pod types are now available in 4 different sizes: 1x, 2x, 4x, and 8x. Capacity and compute per pod double with each size increment.

p2 pod type (Public Preview)("Beta")

The new p2 pod type provides search speeds of around 5ms and throughput of 200 queries per second per replica, or approximately 10x faster speeds and higher throughput than the p1 pod type, depending on your data and network conditions.

This is a public preview feature and is not appropriate for production workloads.

Improved p1 and s1 performance

The s1 and p1 pod types now offer approximately 50% higher query throughput and 50% lower latency, depending on your workload.

July 26, 2022

You can now specify a metadata filter to get results for a subset of the vectors in your index by calling describe_index_stats with a filter object.

The describe_index_stats operation now uses the POST HTTP request type. The filter parameter is only accepted by describe_index_stats calls using the POST request type. Calls to describe_index_stats using the GET request type are now deprecated.

July 12, 2022

Pinecone Console Guided Tour

You can now choose to follow a guided tour in the Pinecone Console. This interactive tutorial walks you through creating your first index, upserting vectors, and querying your data. The purpose of the tour is to show you all the steps you need to start your first project in Pinecone.

June 24, 2022

Updated response codes

The create_index, delete_index, and scale_index operations now use more specific HTTP response codes that describe the type of operation that succeeded.

June 7, 2022

Selective metadata indexing

You can now store more metadata and more unique metadata values! Select which metadata fields you want to index for filtering and which fields you only wish to store and retrieve. When you index metadata fields, you can filter vector search queries using those fields. When you store metadata fields without indexing them, you keep memory utilization low, especially when you have many unique metadata values, and therefore can fit more vectors per pod.

Single-vector queries

You can now specify a single query vector using the vector input. We now encourage all users to query using a single vector rather than a batch of vectors, because batching queries can lead to long response messages and query times, and single queries execute just as fast on the server side.

Query by ID

You can now query your Pinecone index using only the ID for another vector. This is useful when you want to search for the nearest neighbors of a vector that is already stored in Pinecone.

Improved index fullness accuracy

The index fullness metric in describe_index_stats() results is now more accurate.

April 25, 2022

Partial updates (Public Preview)

You can now perform a partial update by ID and individual value pairs. This allows you to update individual metadata fields without having to upsert a matching vector or update all metadata fields at once.

New metrics

Users on all plans can now see metrics for the past one (1) week in the Pinecone console. Users on the Enterprise and Enterprise Dedicated plan now have access to the following metrics via the Prometheus metrics endpoint:

  • pinecone_vector_count
  • pinecone_request_count_total
  • pinecone_request_error_count_total
  • pinecone_request_latency_seconds
  • pinecone_index_fullness (Public Preview)

Note: The accuracy of the pinecone_index_fullness metric is improved. This may result in changes from historic reported values. This metric is in public preview.

Spark Connector

Spark users who want to manage parallel upserts into Pinecone can now use the official Spark connector for Pinecone to upsert their data from a Spark dataframe.

Support for Boolean and float metadata in Pinecone indexes

You can now add Boolean and float64 values to metadata JSON objects associated with a Pinecone index.

New state field in describe_index results

The describe_index operation results now contain a value for state, which describes the state of the index. The possible values for state are Initializing, ScalingUp, ScalingDown, Terminating, and Ready.

Delete by metadata filter

The Delete operation now supports filtering my metadata.