Create an index

Create a Pinecone serverless index for full-text (BM25), semantic (dense vector), sparse-vector, or hybrid search with a document schema.

A Pinecone index lets you store and search your data. A document index stores JSON documents whose fields you declare in a schema at creation. It can combine dense vector, sparse vector, and full-text (string) fields in one document, so a single index serves full-text search, semantic search, and sparse-vector search.

To create a legacy vector index, see Create a vector index. To create a legacy integrated index, see Create an integrated index.

You can create an index using the Pinecone console.

​

Create a document index

An index with a document schema stores typed JSON documents. The schema declares how each ranking field is indexed: as a string field with full_text_search enabled for BM25 ranking, a dense_vector for ANN similarity, or a sparse_vector. A single index can mix all three ranking field types, and at query time you pick the ranking signal with score_by. Any other fields you upsert are stored as metadata, auto-indexed for filtering at upsert time.

Full-text search and integrated embedding are different features. A string field with full_text_search is indexed for BM25 keyword ranking and Lucene queries. It doesn’t create vectors or call an embedding model. To have Pinecone turn a field’s text into vectors, use integrated embedding: declare embed or sparse_embed on a string field in a document index. The legacy create_index_for_model path does this too, creating a legacy integrated index served by the legacy Records API.

Indexes with document schemas use API version 2026-07, supported through REST and the Python (v10 or later), Node.js (v9 or later), and Go (v7 or later) SDKs. For other languages, call the REST endpoint directly.

​

Minimal: BM25 on a single text field

The example below creates an articles index whose body field is indexed for BM25 ranking. Other fields included at upsert time are stored on each document and auto-indexed for filtering as metadata.

Python

import os
from pinecone import Pinecone, SchemaBuilder

pc = Pinecone(api_key=os.environ["PINECONE_API_KEY"])

schema = (
    SchemaBuilder()
      .add_string_field(name="body", full_text_search={})
      .build()
)

pc.indexes.create(name="articles", schema=schema)

​

Multi-field schema: BM25 + dense vector

A single index with a document schema can hold string fields with full_text_search and dense_vector ranking fields together, and the same schema can also include a sparse_vector field. A single search request ranks by one scoring type. Multi-field BM25 is supported: use multiple text clauses on different fields, or one query_string clause spanning fields. Any scoring method can be combined with metadata filters, including text-match filters ($match_phrase, $match_all, $match_any) on string fields with full_text_search.

Python

import os
from pinecone import Pinecone, SchemaBuilder

pc = Pinecone(api_key=os.environ["PINECONE_API_KEY"])

schema = (
    SchemaBuilder()
      .add_string_field(name="title", full_text_search={})
      .add_string_field(name="body", full_text_search={})
      .add_dense_vector_field(name="embedding", dimension=1536, metric="cosine")
      .build()
)

pc.indexes.create(name="articles-multi", schema=schema)

You can include additional fields (for example, category or year) at upsert time. All metadata fields are automatically indexed for filtering — they don’t need to be declared in the schema. The schema is for ranking fields only. string_list, float, and boolean fields are always metadata, and a string field is metadata unless it sets full_text_search, embed, or sparse_embed. Declaring a metadata-only field in the schema is rejected at index creation. To index only the metadata fields you filter on, see Configure metadata indexing.

​

Integrated embedding

On a document index, declare embed or sparse_embed on a string field. Pinecone then embeds its text on write and the query text at search time, so you upsert and search with plain text instead of vectors. See Integrated embedding for the schema, examples, and limits.

For the full schema reference (all field types, language and analyzer options, dedicated read capacity, and Python SDK examples), see Full-text search.

You can’t add, remove, or modify a document index’s fields after you create it. To change a schema, delete the index and create a new one. To move an existing legacy vector index or legacy integrated index onto a document schema, convert it instead of reloading your data.

​

Create an index from a backup

You can create an index from a backup. For more details, see Restore an index.

​

Wait for the index to be ready

A new index of any type can’t accept data until its status.ready field is true. The Python SDK waits for this by default, and the Node.js SDK waits when you pass waitUntilReady: true. If you call the API directly, describe the index repeatedly until status.ready is true, then upsert.