You cannot enable AWS S3 Vector on an existing standard S3 bucket because vector functionality requires a completely different bucket type called a “vector bucket.” Vector buckets are purpose-built storage containers that use different APIs, data structures, and optimization strategies compared to standard S3 buckets. To use S3 Vector capabilities, you must create a new vector bucket specifically designed for storing and querying vector data. This architectural decision ensures optimal performance for vector operations while maintaining the reliability and familiarity of standard S3 for traditional file storage needs.
Creating a vector bucket involves using either the AWS Management Console, AWS CLI, or AWS SDKs with the new s3vectors service namespace. In the console, you navigate to the “Vector buckets” section (separate from standard S3 buckets), click “Create vector bucket,” and specify a unique bucket name following the naming requirements (3-63 characters, lowercase letters, numbers, and hyphens only). You must also choose encryption settings during creation - either server-side encryption with S3-managed keys (SSE-S3) or customer-managed KMS keys (SSE-KMS). These encryption settings cannot be changed after bucket creation, making proper planning essential.
After creating your vector bucket, you need to create vector indexes within it to organize your vector data. Each vector index requires configuration of dimension size (matching your embedding model’s output), distance metric for similarity calculations, and optional non-filterable metadata keys. You can create multiple indexes within a single vector bucket (up to 10,000 per bucket) to organize different types of vector data or support different applications. If you have existing vector data in standard S3 buckets or other storage systems, you’ll need to extract that data, potentially reprocess it into the proper vector format, and ingest it into your new vector indexes using the PutVectors API operation.
Will Amazon S3 vectors kill vector databases or save them?
S3 vectors looks great particularly in terms of price and integration into the AWS ecosystem. So naturally, there are a lot of hot takes. I’ve seen folks on social media and in engineering circles say this could be the end of purpose-built vector databases—Milvus, Pinecone, Qdrant, and others included. Bold claim, right?
As a group of people who’s spent way too many late nights thinking about vector search, we have to admit that: S3 Vectors does bring something interesting to the table, especially around cost and integration within the AWS ecosystem. But instead of “killing” vector databases, I see it fitting into the ecosystem as a complementary piece. In fact, its real future probably lies in working with professional vector databases, not replacing them.
Check out James’ post to learn why we think that—looking at it from three angles: the tech itself, what it can and can’t do, and what it means for the market. We’ll also share S3 vectors’ strenghs and weakness and in what situations you should choose an alternative such as Milvus and Zilliz Cloud.
Will Amazon S3 Vectors Kill Vector Databases—or Save Them?
Or if you’d like to compare Amazon S3 vectors with other specialized vector databases, visit our comparison page for more details: Vector Database Comparison