--- title: Pinecone API Integration on Truto slug: pinecone category: Database canonical: "https://truto.one/integrations/detail/pinecone/" --- # Pinecone API Integration on Truto **Category:** Database **Status:** Generally available ## MCP-ready AI tools Truto exposes 54 tools for Pinecone that AI agents can call directly. - **get_single_pinecone_index_by_id** — Get a single Pinecone index by id. Returns the index configuration including name, status, host, dimension, metric, spec, vector_type, deletion_protection, and tags. Required: id. - **create_a_pinecone_index** — Create a new Pinecone index for dense vectors, sparse vectors, or full-text search with a document schema. Returns the created index object including name, host, status, and spec. Required: name and spec.serverless (cloud and region) for standard indexes, or deployment (cloud and region) for document-schema indexes. - **list_all_pinecone_indexes** — List all Pinecone indexes in the current project. Returns an array of index objects each including name, dimension, metric, status, host, and spec. - **update_a_pinecone_index_by_id** — Update the configuration of a Pinecone index by id. Returns the updated index object. Required: id. - **delete_a_pinecone_index_by_id** — Delete a Pinecone index by id. Returns an empty response on success. Required: id. - **pinecone_indexes_search** — Search Pinecone indexes. Returns matching index objects. - **update_a_pinecone_vector_by_id** — Update the values and/or metadata of a single vector in a Pinecone index. Required: `id`. - **delete_a_pinecone_vector_by_id** — Delete vectors from a Pinecone index by IDs or a metadata filter expression. Returns an empty response on success. - **pinecone_vectors_bulk_create** — Upsert (insert or overwrite) a batch of dense or sparse vectors into a Pinecone index namespace. Returns the count of upserted vectors. Required: `vectors`. - **get_single_pinecone_vector_by_id** — Fetch one or more vectors from a Pinecone index by their IDs. Returns each vector's id, values, and metadata. Required: `ids`. - **list_all_pinecone_vectors** — List vector IDs stored in a Pinecone index namespace. Returns a list of vector id strings, optionally filtered by ID prefix. - **pinecone_vectors_search** — Query a Pinecone index for the nearest matching vectors to a given query vector or existing vector ID. Returns matches including id and score. Required: `topK`. - **delete_a_pinecone_chunk_by_id** — Delete chunks from a Pinecone index by namespace and filter criteria. Supports filtering by fields such as document_id and chunk_number to target specific chunks for removal. Required: namespace and filter. - **create_a_pinecone_import** — Start an asynchronous bulk import of vectors from object storage (S3, GCS, or Azure Blob Storage) into a Pinecone serverless index. Returns the `id` of the import operation for subsequent status tracking. Required: `uri`. - **delete_a_pinecone_import_by_id** — Cancel a Pinecone bulk import operation by id if it has not yet finished. Has no effect if the import is already complete. Returns an empty response on success. Required: `id`. - **get_single_pinecone_import_by_id** — Get details of a specific Pinecone bulk import operation by id. Returns: id, uri, status, percent_complete, records_imported, created_at, finished_at, and error (if failed). Required: `id`. - **list_all_pinecone_imports** — List all recent and ongoing Pinecone bulk import operations for an index. Returns: id, uri, status, percent_complete, records_imported for each import. - **create_a_pinecone_collection** — Create a Pinecone collection (static backup) from a pod-based index. Returns the created collection object including name and status. Required: name, source. Only supported for pod-based indexes; serverless indexes do not support collections. - **list_all_pinecone_collections** — List all Pinecone collections in the current project. Returns an array of collection objects including name, status, environment, size, vector_count, and dimension. Serverless indexes do not support collections. - **delete_a_pinecone_record_by_id** — Delete records from a Pinecone namespace by ID list, metadata filter, or all records at once. Returns an empty response on success. Required: `namespace`. Provide `ids` (up to 1000), a metadata `filter` expression, or `deleteAll: true` in the request body. - **list_all_pinecone_records** — List records in a Pinecone namespace. Returns an array of record objects including id and values. Required: `namespace`. - **pinecone_records_search** — Search a Pinecone namespace for semantically similar records using a text query. Returns hits with `_id`, `_score`, and requested `fields`. Required: `namespace` and `query` body object containing `top_k` and `inputs.text`. - **create_a_pinecone_record** — Upsert records into a Pinecone namespace using NDJSON format; inserts new records or updates existing ones matched by `id`. Returns an empty response on success. Required: `namespace`. Body must be newline-delimited JSON with one record object per line. - **create_a_pinecone_project** — Create a new Pinecone project in your organization. Returns: id, name, organization_id, created_at, max_pods, force_encryption_with_cmek. Required: name. - **delete_a_pinecone_project_by_id** — Delete a Pinecone project and all its associated configuration by id. All indexes, assistants, backups, and collections within the project must be removed before deletion; API keys are deleted automatically. Required: id. - **get_single_pinecone_project_by_id** — Get details about a specific Pinecone project by id. Returns: id, name, max_pods, force_encryption_with_cmek, organization_id, created_at. Required: id. - **list_all_pinecone_projects** — List all Pinecone projects in your organization. Returns an array of project objects each including id, name, max_pods, force_encryption_with_cmek, organization_id, and created_at. - **update_a_pinecone_project_by_id** — Update a Pinecone project's configuration by id. Supports changing the project name, maximum Pod count, or CMEK encryption setting. Returns the updated project including id, name, max_pods, force_encryption_with_cmek, organization_id, and created_at. Required: id. - **pinecone_documents_bulk_create** — Upsert documents into a Pinecone index namespace, inserting new documents or fully replacing existing ones matched by _id. Returns upserted_count. Required: namespace and documents array (1–1000 items, each with _id). Documents are indexed asynchronously and may not be immediately searchable. - **pinecone_documents_search** — Search Pinecone documents in a namespace using BM25 text, Lucene query_string, dense vector, or sparse vector ranking, with optional pre-scoring filters. Returns matches array with _id, _score, and any requested fields. Required: namespace, score_by, top_k. Max top_k is 10,000. - **list_all_pinecone_search** — Search a Pinecone index for records using a text query, sparse vector, or record ID. Returns hits with _id, _score, and matched field values. Required: namespace. - **create_a_pinecone_api_key** — Create a new Pinecone API key for a project. Returns the key object including id, name, project_id, roles, and the plaintext key value. Required: project_id, name, roles. - **delete_a_pinecone_api_key_by_id** — Delete a Pinecone API key by id. Returns an empty response on success. Required: id. - **get_single_pinecone_api_key_by_id** — Get details of a Pinecone API key by id, excluding the secret value. Returns: id, name, project_id, roles. Required: id. - **list_all_pinecone_api_keys** — List all Pinecone API keys in a project. Returns: id, name, project_id, roles for each key. Required: project_id. - **update_a_pinecone_api_key_by_id** — Update the name and roles of a Pinecone API key by id. Returns the updated key object including id, name, project_id, roles, and key value. Required: id. - **create_a_pinecone_backup** — Create a backup of a Pinecone index. Returns the backup object including backup_id, status, cloud, region, and created_at. Required: index_name. - **delete_a_pinecone_backup_by_id** — Delete a Pinecone backup by id. Returns an empty response on success. Required: id. - **get_single_pinecone_backup_by_id** — Get details of a Pinecone backup by id. Returns: backup_id, source_index_name, status, record_count, namespace_count, and created_at. Required: id. - **list_all_pinecone_backups** — List all backups for a Pinecone project. Returns an array of backup objects each including backup_id, source_index_name, status, record_count, and created_at. - **get_single_pinecone_restore_job_by_id** — Get a description of a specific Pinecone restore job by id. Returns: restore_job_id, backup_id, target_index_name, status, created_at, completed_at, percent_complete. Required: id. - **list_all_pinecone_restore_jobs** — List all Pinecone restore jobs for a project. Returns an array of restore job objects including restore_job_id, backup_id, target_index_name, status, created_at, completed_at, and percent_complete. - **list_all_pinecone_index_backups** — List all backups for a Pinecone index. Returns: backup_id, name, status, cloud, region, and created_at for each backup. Required: index_name. - **create_a_pinecone_namespace** — Create a namespace in a Pinecone serverless index. Returns the created namespace including name, record_count, and schema. Required: name. Not supported for pod-based indexes. - **delete_a_pinecone_namespace_by_id** — Delete a namespace from a Pinecone serverless index by id. This operation is irreversible and permanently removes all data in the namespace. Returns an empty response on success. Required: id. Not supported for pod-based indexes. - **get_single_pinecone_namespace_by_id** — Get details about a specific namespace in a Pinecone serverless index by id. Returns: name, record_count. Required: id. Not supported for pod-based indexes. - **list_all_pinecone_namespaces** — List all namespaces in a Pinecone serverless index in sorted order. Returns an array of namespaces each with name and record_count. Not supported for pod-based indexes. - **get_single_pinecone_model_by_id** — Get a description of a specific Pinecone model by id. Returns: model name, type, vector_type, default_dimension, modality, provider_name, supported_metrics, supported_dimensions, and supported_parameters. Required: id. - **list_all_pinecone_models** — List all embedding and reranking models hosted by Pinecone. Returns an array of model objects each including model name, type, vector_type, default_dimension, modality, provider_name, and supported_parameters. - **create_a_pinecone_embedding** — Generate vector embeddings for input data using Pinecone's hosted embedding models. Returns: data (array of embedding value vectors), model name, and usage stats including total_tokens. Required: model and inputs. - **create_a_pinecone_rerank** — Rerank a list of documents by relevance to a query using Pinecone's inference API. Returns a ranked data array with each result's index, document content, and relevance score. Required: model, query, and documents. - **create_a_pinecone_integrated_index** — Create a Pinecone integrated index pre-configured for a specific embedding model. Returns the created index object. Required parameters must be supplied in the request body. - **list_all_pinecone_vector_records** — List vector records from a Pinecone index by fetching them via metadata filtering. Returns a collection of vectors stored at the resolved index host. Required: index_host. - **get_single_pinecone_index_stat_by_id** — Get statistics for a Pinecone index via the describe_index_stats endpoint. Returns index-level metrics such as total vector count and per-namespace breakdowns. ## How it works 1. **Link your customer's Pinecone account.** Use Truto's frontend SDK; we handle every OAuth and API key flow so you don't need to create the OAuth app. 2. **Authentication is automatic.** Truto refreshes tokens, stores credentials securely, and injects them into every API request. 3. **Call Truto's API to reach Pinecone.** The Proxy API is a 1-to-1 mapping of the Pinecone API. 4. **Get a unified response format.** Every response uses a single shape, with cursor-based pagination and data in the `result` field. ## Use cases - **Embed RAG into your SaaS product** — Let your customers connect their own Pinecone account so your AI features (chatbots, copilots, smart search) retrieve context from their proprietary data without you hosting their vectors. - **Offer a Pinecone destination for data sync** — If you run an ETL, CMS, helpdesk, or CRM, let users automatically push their content into Pinecone as vectors so it becomes searchable by their AI stack. - **Multi-tenant vector isolation for your users** — Provision a dedicated Pinecone namespace or index per end-customer directly from your app, giving each tenant isolated vector storage without manual setup. - **Compliance-ready data lifecycle management** — Automatically delete vectors, records, or chunks from a customer's Pinecone index when they remove the source document in your app, keeping GDPR and audit workflows clean. - **No-code AI workflow building blocks** — Expose Pinecone as a step in your visual workflow or automation builder so non-technical users can embed text, upsert vectors, and run semantic searches without writing code. ## What you can build - **Index picker for end-user onboarding** — Use list_all_pinecone_indexes and get_single_pinecone_index_by_id to let users pick or auto-provision the index your app should read from and write to. - **Continuous knowledge sync to Pinecone** — Push incremental updates from your app into the user's index using pinecone_vectors_bulk_create or pinecone_documents_bulk_create whenever source content changes. - **In-app semantic search and RAG retrieval** — Power AI responses by calling pinecone_vectors_search, pinecone_documents_search, or pinecone_records_search with metadata filters scoped to the active user or workspace. - **Native embedding and reranking pipeline** — Use create_a_pinecone_embedding to vectorize text on the fly and create_a_pinecone_rerank to reorder retrieved results before passing them to your LLM. - **Namespace-per-tenant provisioning** — Automatically call create_a_pinecone_namespace when a new workspace, project, or customer is created, and clean up with delete_a_pinecone_namespace_by_id on offboarding. - **Bulk historical import from object storage** — Trigger create_a_pinecone_import for large initial syncs from S3/GCS and monitor progress via list_all_pinecone_imports and get_single_pinecone_import_by_id. ## FAQs ### How do end users authenticate their Pinecone account? Pinecone uses API key authentication. Through Truto, your users provide their Pinecone API key during the connect flow, and Truto securely stores and injects it into every request — you never have to handle the credential yourself. ### Can we provision isolated storage per end-customer? Yes. You can either create a dedicated index per tenant with create_a_pinecone_index, or use a shared index and isolate tenants via create_a_pinecone_namespace. Truto exposes both lifecycle operations including delete and list endpoints. ### Does Truto support Pinecone's serverless inference for embeddings and reranking? Yes. create_a_pinecone_embedding generates vectors from raw text using Pinecone's hosted models, and create_a_pinecone_rerank reorders search results by relevance — both are accessible through the same unified connection. ### How do we keep vectors in sync when source data changes or is deleted? Use pinecone_vectors_bulk_create or pinecone_documents_bulk_create for upserts, and delete_a_pinecone_vector_by_id, delete_a_pinecone_record_by_id, or delete_a_pinecone_chunk_by_id to remove stale data — typically triggered from your app's own create/update/delete webhooks. ### Can we run hybrid search combining keyword and vector queries? Yes. pinecone_documents_search supports Pinecone's document schema which combines BM25 keyword scoring with vector similarity, and pinecone_vectors_search supports dense, sparse, and metadata-filtered queries. ### How do we handle very large initial data loads for a new customer? Use create_a_pinecone_import to ingest large datasets directly from object storage (S3/GCS) asynchronously, then poll get_single_pinecone_import_by_id or list_all_pinecone_imports to track completion before switching to incremental bulk upserts.