# From 48 Seconds to 130 Milliseconds: Vector Search in Tinybird

DevFeed: [From 48 Seconds to 130 Milliseconds: Vector Search in Tinybird](<https://devfeed.tech/articles/from-48-seconds-to-130-milliseconds-vector-search-in-tinybird-18739.md>)

Original publisher: [Read original article](<https://www.tinybird.co/blog/vector-search-improvements>)

Author: Daniel Sangorrín

Published: 2026-05-12T00:00:00Z

Content type: article

Language: en

Sources: [Tinybird](<https://devfeed.tech/sources/tinybird.md>)

Topics: [Embeddings](<https://devfeed.tech/topics/embeddings.md>), [AI search](<https://devfeed.tech/topics/ai-search.md>)

Tags: [customer](<https://devfeed.tech/tags/customer.md>), [embeddings](<https://devfeed.tech/tags/embeddings.md>), [engineering-excellence](<https://devfeed.tech/tags/engineering-excellence.md>), [search](<https://devfeed.tech/tags/search.md>), [vector](<https://devfeed.tech/tags/vector.md>), [vector-search](<https://devfeed.tech/tags/vector-search.md>)

## AI overview

A customer needed semantic search over 20 million embeddings, but the initial attempt timed out. The article reports that three changes reduced query latency to under 200 milliseconds.

## Source excerpt

A customer needed semantic search over 20 million embeddings. Their first attempt timed out. Three changes turned it into sub-200ms queries. Here's what we learned.