Chroma Vector Database

AI Platform & Data chroma 1.5.9 Hardened AMI

Chroma 1.5.9 on Ubuntu 24.04 stores embeddings you supply and queries them through a JSON API. One collection lives on local disk. The walkthrough writes three sample vectors and a separate reader checks them. The image does not download an embedding model, and it does not include a browser UI.

Get started View on AWS Marketplace

AMI release: 0.1.0 (chroma 1.5.9)
Recommended: t3.medium
Storage: 32 GiB+ encrypted gp3
Access: Session Manager or SSH
Released: Oct 25, 2026
Delivery: Amazon Machine Image in your own AWS account

Persistent Chroma vector database API for your own embeddings, with data kept on local storage.

Listens on loopback only; reach it through a Session Manager or SSH tunnel (no browser UI included).

Included demo writes and queries sample vectors, and a separate check verifies they persist.

Quick start

Full guide →

What is on the image¶

This guide matches Marketplace package 0.1.0 (chroma 1.5.9). Check the listing for the version currently on offer.

Chroma 1.5.9, the chroma service, running as the chroma service account.
A JSON API on 127.0.0.1:8000. There is no bundled browser dashboard.
Local storage in /var/lib/chroma and configuration in /etc/chroma/config.yaml.
A reader, sushi-chroma read, and a walkthrough at /opt/sushi/docs/demo.sh.
Notices under /opt/chroma and /usr/share/doc/chroma. Installed inventory is under /opt/sushi/evidence.

Reset is disabled. CORS origins are empty. Requests are limited to 1 MiB. Telemetry export is disabled. The service network is limited to loopback. First boot creates storage once and then keeps the instance identity.

Recommended instance¶

t3.medium (4 GiB of memory) with an encrypted gp3 root volume of at least 32 GiB and IMDSv2. Leave inbound application ports closed. Reach port 8000 through Session Manager or an SSH tunnel.

Billing¶

The listing states a software rate of USD 0.03 per running instance-hour. Confirm the offer before you subscribe. EC2, EBS, and data transfer are billed separately. No embedding-provider key is required, because you send the vectors yourself.

Tutorial¶

Watch on YouTube.

Where to go next¶

Getting started: launch, run the demo, and query the saved vectors.
Configuration: listener, request size, and embeddings.
Operations: service, backups, and support.
Troubleshooting: startup and a rejected listener change.

Upstream project¶

Chroma is an open source vector database. Software Sushi packages this image independently and does not claim upstream or AWS endorsement. Delivered notices are under /opt/chroma and /usr/share/doc/chroma.

The Chroma documentation describes the upstream project. This server does not implement application authentication. Do not treat legacy CHROMA_SERVER_AUTHN variables as protection for this build. Access is the instance login and the tunnel.

Last updated October 27, 2026 · Suggest an improvement