Free during the pilot

Test Asteroid on your AI workload

See whether a memory-friendly vector database can lower your AI infrastructure cost. We provide a pilot instance and help you get your first data loaded.

Request a pilot

What you get

A dedicated instance

A dedicated Asteroid instance on AWS, reachable over HTTPS with an API key, ready to test with your data.

Client + HTTP API

Use the Python client or call the REST API from any language.

Hands-on help

We help you load data, run first queries, and understand fit for your workload.

Free for the pilot

No cost, no card, no self-serve signup. We onboard each pilot directly.

How to apply

1

Tell us about your workload

Tell us your use case and rough scale: vector count, dimension, and expected traffic.

2

We provision your instance

We provision an instance for your dimension and metric, then send the endpoint and API key.

3

Load data and run queries

Insert your vectors, run searches, and see how Asteroid performs on your workload.

Pilot pricing

The pilot is free. After the pilot, pricing will be flat monthly with fair-use limits: no per-query, per-read, or per-vector meter. Tell us your use case and we will suggest the right plan.

Ready to spend less on vector search?

Request a pilot