The Postgres development platform. Supabase gives you a dedicated Postgres database to build your web, mobile, and AI applications.
TypeScript aialternativeauthdatabasedenoembeddingsexamplefirebasenextjsoauth2pgvectorpostgispostgrespostgresqlpostgrestrealtimesupabasevectorswebsockets
---
title: 'Vector Bucket Limits'
subtitle: 'Understanding capacity, limits, and billing for vector buckets.'
---
<Admonition type="caution" title="This feature is in alpha">
Expect rapid changes, limited features, and possible breaking updates. [Share feedback](https://github.com/orgs/supabase/discussions/40116) as we refine the experience and expand access.
</Admonition>
Vector buckets have default limits during the alpha phase. These limits are designed to ensure fair resource allocation and can be adjusted on a case-by-case basis for production workloads.
## Storage limits
| Limit | Value | Notes |
| ----------------------- | ------------ | ----------------------------------------------------- |
| **Buckets per project** | 10 | Maximum number of vector buckets per Supabase project |
| **Indexes per bucket** | 10 | Maximum number of indexes per bucket |
| **Vector dimensions** | Max 4096 | Maximum dimension size for embeddings |
| **Batch size** | 1000 vectors | Maximum vectors per single insert/update request |
## Query limits
The `queryVectors` limits differ between hosted and local vector buckets.
| Limit | Hosted vector buckets | Local vector buckets |
| -------------------------------- | --------------------- | -------------------- |
| **Maximum `topK`** | 10,000 | 100 |
| **Maximum results per response** | 100 | 100 |
| **`nextToken` query pagination** | Supported | Not supported |
Hosted vector buckets use Amazon S3 Vectors. When a response includes a `nextToken`, send the same query again with that token to retrieve the next page. Local vector buckets use pgvector and return the requested results in one response.