No fuss vector database for AI

No fuss means: no resource calculations, no pod sizes, no complicated schema definitions, no installation, no complex API, low-cost and ready-to-use.

SemaDB Cloud

Get started with SemaDB using our hosted solution via RapidAPI which comes with an easy-to-use RESTful API and an interactive playground.

RESTful API

Insert and search for points using a familiar JSON API. No custom clients or initialisation required. The entire API is documented online with examples and tutorials.

Low cost

For 1536 dimensional vectors, SemaDB Cloud offers one of the lowest cost options for getting started with vector similarity search.

With a generous free tier that doesn't expire, you can go live at no cost.

Pick your language

SemaDB doesn't require a custom client to interactive with. Instead the RESTful API can be used from any language you prefer with examples for each one for each endpoint.

Head over to RapidAPI to see working examples of how to use SemaDB with your favourite language.

On-Site Hosting

You can integrate SemaDB into your search stack with various installation options by hosting it yourself.

Consulting

We can help you build and deliver an AI solution around vector search using our products. Contact us

What indexing algorithm does SemaDB use?

SemaDB uses a variant of the disk-based Vamana algorithm to store and index points in a collection. This reduces the dependency on RAM and resource requirements. For most applications the increase in latency is worth it.

Where is SemaDB Cloud hosted?

We are based in the United Kingdom and our servers for SemaDB Cloud are located in London. In the future we may offer multiple regions to choose from.