AI Search pricing & limits

Question

I'm struggling to understand the quotas / limits / tiers of Azure AI Search.

This is what's shown for the pricing:

I have a Basic search service, with 5 indexes and this is what I see under usage:

Where does that 2GB limit and 1GB limit come from? Where is my 15GB limit? (And why is my Vector index size 0 Bytes?)

If I go to Scale, then I can't increase the partitions because they only apply to Standard tiers:

Answer 1

Hello Sean !

Thank you for posting on Microsoft Learn.

From your second image, Total storage (Quota: 2 GB): This is not the full 15 GB partition quota listed in the pricing table.

In the Basic tier, the 15 GB is the maximum per partition, but your current setup has only 1 partition and 1 replica, resulting in a 2 GB limit shown in the portal.

Microsoft reserves some storage for system use, indexing overhead, etc., which is why it shows 2 GB user-usable storage rather than the full 15 GB.

Vector index size (Quota: 1 GB): This is your vector storage quota under the Basic tier.

Since you're not using vector search features (no embeddings created), it's showing 0 Bytes used.

Indexes (Quota: 15): You're using 5 out of your 15 allowed indexes in the Basic tier.

Where is the 15 GB Storage per Partition?

Your Basic SKU provides 15 GB per partition, but Azure AI Search in Basic limits you to a single partition, so you can’t scale out horizontally like in Standard or Storage Optimized tiers.

The 2 GB limit is likely what is available for actual document storage after accounting for:

System overhead

Inverted index structures

Reserved metadata space

This is a displayed quota, not the raw 15 GB allocation. You cannot increase the number of partitions in the Basic tier only Standard and above support that.

Why can’t you increase partitions?

From image 3, you are on the Basic tier, which supports only 1 partition.

Partition scaling is only available in Standard (S, S2, S3, etc.) and Storage Optimized (L1, L2) tiers.

So when you try to scale partitions, the UI blocks that because you're not on a tier that supports it.

Why Is Vector Index Size 0 Bytes?

You're not using vector capabilities yet. Azure AI Search only starts allocating vector index storage when:

You create a vector field in an index

And upload vector data (like embeddings)

Your current usage suggests no vector data has been ingested yet.