Bring your own vectors

If you already generate embeddings, store them directly. Create an index with a dense-vector field, upsert your vectors, and rank by similarity. Pinecone does no embedding on your behalf here.

Prerequisites

A Pinecone account and API key (get one).
Python 3.10+.
The Pinecone Python SDK: pip install --upgrade pinecone.

Upsert your vectors and search

Set your API key

Set your API key as an environment variable so the SDK can authenticate:

export PINECONE_API_KEY="YOUR_API_KEY"

Create an index with a dense-vector field

Set dimension to match your embedding model’s output, and pick a distance metric (cosine, dotproduct, or euclidean). Only the vector field goes in the schema; other fields like text are stored on the documents (non-schema fields are stored as metadata, capped at 40 KB per document).

import os, time
from pinecone import Pinecone, SchemaBuilder

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

schema = (
    SchemaBuilder()
      .add_dense_vector_field(name="embedding", dimension=1024, metric="cosine")
      .build()
)
if not pc.indexes.exists(name="byo-vectors"):
    pc.indexes.create(name="byo-vectors", schema=schema)

while not pc.indexes.describe(name="byo-vectors").status.ready:
    time.sleep(2)

index = pc.index(name="byo-vectors")

Upsert your vectors

Each document carries its embedding (a list of floats matching the schema’s dimension) plus any fields you want to store. The vectors below are stand-ins with the right dimension. Replace them with your real embeddings.

index.documents.upsert(
    namespace="__default__",
    documents=[
        {"_id": "1", "embedding": [0.1, 0.2] * 512, "text": "Refund requests must be submitted within 30 days."},
        {"_id": "2", "embedding": [0.2, 0.1] * 512, "text": "Enterprise support responds within 4 hours."},
    ],
)

time.sleep(5)  # documents are indexed asynchronously, so wait a moment

Search by vector similarity

Embed your query with the same model you used for the documents, then rank by the dense-vector field (embedding).

query_embedding = [0.1, 0.2] * 512  # replace with your query's embedding from the same model

resp = index.documents.search(
    namespace="__default__",
    top_k=3,
    score_by=[{"type": "dense_vector", "fields": ["embedding"], "values": query_embedding}],
    include_fields=["*"],
)

for m in resp.matches:
    print(m._id, m._score, getattr(m, "text", ""))

Next steps

Match keywords

Add a full_text_search field to your schema for keyword search, or combine it with your vectors for hybrid search.

Data modeling

How to design a schema for your workload

Get started

Indexes

Data

Search

Optimize

Cost

Production

Agents

Pod-based indexes (legacy)

Prerequisites

Upsert your vectors and search

Next steps

Match keywords

Data modeling

​Prerequisites

​Upsert your vectors and search

​Next steps

Match keywords

Data modeling

Prerequisites

Upsert your vectors and search

Next steps