Skip to main content
Welcome — in this lesson you’ll learn how to create an AWS S3 vector bucket via the AWS Management Console and how to create a vector index inside that bucket. The steps below follow the console flow and include best practices for naming, dimensions, and similarity metrics.

​
Overview

Steps covered:
  1. Create an S3 vector bucket.
  2. Create a vector index inside the bucket.
  3. Review encryption and access settings.
  4. Notes on vector dimensions and programmatic ingestion.

​
1) Create the S3 vector bucket

  1. Sign in to the AWS Management Console.
  2. From the console home, search for “S3” and open the Amazon S3 service.
  3. In the S3 left-hand menu, select “Vector buckets”.
  4. Click Create vector bucket, provide a globally unique name, and configure encryption, tags, or access controls as needed.
Bucket naming rules to remember:
  • Lowercase letters, numbers, and hyphens only
  • Length between 3 and 63 characters
  • Must be globally unique
The image shows an AWS console interface where a user is creating a vector bucket named "airline-policy-vectors-raghu" with options for encryption and tagging.
Scroll down, accept defaults (or set required encryption/tags/access controls for your environment), then click Create bucket. The bucket will be created and listed in your S3 console.

​
2) Create a vector index inside the bucket

  1. Open the vector bucket you just created.
  2. Click Create vector index.
  3. Provide a name for the index (e.g., airline-policy-index).
  4. Specify the vector dimension — the length of the embedding vectors you will store (for example, 3, 768, or 1536 depending on your embedding model).
  5. Select a distance metric for similarity search (e.g., cosine similarity or Euclidean distance). For this demo we use cosine similarity.
The image shows an AWS console interface for creating a vector index, with fields for setting the vector index name, dimension, and distance metric. It includes options for cosine and Euclidean distance metrics.
Set the vector dimension to match the output dimension of the embedding model you plan to use (for example, 1536, 768, etc.). If the dimensions don’t match, embeddings cannot be indexed or queried correctly.
Leave the remaining settings at their defaults unless you require custom encryption keys, special access policies, or specific performance/partitioning behavior. Review encryption and permissions before creating the index.
The image shows an AWS console screen focused on setting encryption options for a vector index in Amazon S3, with server-side encryption settings being specified.

​
3) Verify the index

After creation, the vector index is listed inside the bucket. The index uses the selected similarity metric and the dimension you provided.
The image displays an Amazon S3 console screen showing the creation of a vector index named "airline-policy-index" within the "airline-policy-vectors-raghu" bucket. There are tabs for vector indexes, properties, and permissions.

​
Quick reference table

SettingWhat it controlsRecommendation / Example
Bucket nameGlobal identifier for the vector bucketUse lowercase, numbers, hyphens; e.g., airline-policy-vectors-raghu
Vector index nameLogical name for the index inside the bucketairline-policy-index
Vector dimensionLength of embedding vectorsMatch your model output (e.g., 1536, 768)
Distance metricSimilarity measure used for nearest-neighbor searchcosine or Euclidean; use cosine for directional similarity

​
4) Next steps — programmatic ingestion

You can connect to the S3 vector bucket programmatically to ingest embeddings into the index and perform similarity searches. Typical workflow:
  • Generate embeddings using your chosen model.
  • Ensure embeddings have the same dimension as the index.
  • Use the AWS SDK or supported APIs to upload vectors and metadata to the index.
  • Run similarity queries (e.g., k-NN using the chosen metric) to retrieve nearest neighbors.
If you want, I can provide sample code snippets for embedding generation and programmatic ingestion using a specific SDK (Python boto3, JavaScript AWS SDK v3, etc.).

Watch Video