CatBench Vector Search Playground

submited by

Style Pass

2025-01-27 04:00:03

Last year, I published a simple demo app that visualizes embedding vectors of a bunch of cat, dog and airplane photos as heatmaps. It’s a simple concept and does not have much practical use. I built it mostly for fun, but also for some “learning by doing” too.

With this new app, you can interactively navigate through multiple Postgres-based vector similarity search examples and immediately see which SQL statements were executed in the backend and their execution plans with performance metrics. All this is visible right in the web app itself as you explore around. No need to switch to searching through Postgres logs or manually rerunning explain plan for some query in a different tool.

For example, if you pick a cat image from the web app gallery, it performs a vector similarity search using a Postgres SQL query. In addition to the application output, CatBench shows you both the SQL queries used for generating this result and their execution plans with performance metrics:

This makes exploring the performance characteristics of such application queries easier, as you can just navigate around or refresh the current browser page after making some changes, like changing the SQL text or creating a different (vector) index.