Chroma
Open-source embedding database for vector, full-text and metadata search in RAG apps.
Quick answer
Chroma is an open-source embedding database for vector, full-text and metadata search in RAG apps, listed under Data labeling & vector databases. A typical small-business use: vector store for an internal RAG chatbot or document search. Official hardware: minimum 2 GB RAM. That fits a Hetzner Cloud CPX12 server at about €11.99/month (server only, excl. VAT). License: Apache-2.0, which allows internal business use.
Last updated · facts verified against official docs on · sources
Activity data comes from GitHub and was last refreshed Oct 23, 2026.
What is Chroma?
Chroma stores documents, embeddings and metadata in collections and answers similarity queries with metadata and full-text filters. It can run embedded inside a Python or JavaScript app, or as a single-node server (one Docker container) that clients reach over HTTP. Many RAG and agent frameworks support it as a vector store. A hosted Chroma Cloud exists, but the server is Apache-2.0 open source.
Where it fits in a 10–200 person business
- Vector store for an internal RAG chatbot or document search
- Embedded prototype store that later moves to a self-hosted Chroma server
- Semantic search over support tickets, products or knowledge-base articles
These use cases are our editorial suggestions, not claims made by the project.
Can a business use Chroma? (license)
| License | Apache-2.0 |
|---|---|
| Type | Permissive: MIT / Apache-2.0: use, modify and sell with few obligations. |
| Internal business use | Allowed under the license text |
| What to watch | Permissive. Internal and commercial use allowed; keep license and notices. |
This is a plain-English summary of the license text, not legal advice. Third-party models you run with it have their own licenses.
Data & compliance: does Chroma phone home?
| Telemetry (usage data sent to the vendor) | None (documented) As of version 1.5.4, Chroma no longer collects product telemetry. You can still send OpenTelemetry traces to your own collector. |
|---|---|
| Other calls home by default (updates, marketplaces, default cloud services) | Not documented |
| Single sign-on (OIDC / SAML / LDAP) | Not supported By protocol: OIDC not supported · SAML not supported · LDAP not supported. Since v1.0.0 the self-hosted server ships no built-in authentication at all, so there is no login or SSO; add access control in front of it. Source: github.com/chroma-core/chroma/blob/main/docs/mintlify/reference/server-env-vars.mdx |
| Audit log | Not documented |
| Role-based access control (RBAC) | No No built-in authentication or authorization implementation since v1.0.0 (the old authz settings are legacy). Source: github.com/chroma-core/chroma/blob/main/docs/mintlify/reference/server-env-vars.mdx |
| Data residency / GDPR notes from the project | Not documented |
Checked against the project's own docs and repository on . "Not documented" means we could not find the answer in official sources; it does not mean the feature is missing or that no data is sent. Third-party services you connect (model APIs, vector databases, search engines) have their own data policies. Compare all apps.
What hardware does Chroma need?
| Minimum | 2 GB RAM |
|---|---|
| Recommended | Not published |
| GPU | CPU-only; the single-node sizing guide is based on CPU instances. |
The docs say deploying Chroma with less than 2 GB of RAM is not recommended. The HNSW index must fit in RAM: for 1024-dimensional embeddings, max collection size in millions is about 0.245 x RAM in GB, plus at least 1 GB for the system. Disk should be at least as large as RAM plus several GB.
How much does it cost to host Chroma per month?
| Minimum spec | 2 GB RAM → CPX12 (1 vCPU, 2 GB, 40 GB disk) €11.99/mo |
|---|
How we got this: we took the cheapest Hetzner Cloud Regular Performance (shared AMD vCPU) plan with at least the RAM and vCPU the project lists, using prices read from www.hetzner.com/cloud/regular-performance on 2026-10-22. "From" monthly price cap shown by Hetzner, excluding VAT. Prices vary by location. Covers the server only; it excludes LLM API usage, backups, extra storage, GPUs and your time. Disk is not matched; add a volume if you need more than the plan includes.
How to deploy Chroma
Method: Docker
docker run -v ./chroma-data:/data -p 8000:8000 chromadb/chroma
# clients connect with HttpClient(host='localhost', port=8000)From the official instructions: docs.trychroma.com/guides/deploy/docker. Always check the source for the current version before you run it.
Known gotchas when self-hosting Chroma
- Chroma v1.0.0 and later no longer ship built-in authentication, and the server binds to 0.0.0.0 by default. Keep port 8000 on a private network or behind your own authenticating proxy. source
- If a collection grows larger than available RAM, latency spikes as the OS swaps and the system quickly becomes unusable. Size RAM for the total number of embeddings you expect. source
- Single-node Chroma does not scale forever; the docs suggest a distributed deployment when you exceed the single-node limits. source
Alternatives in Data labeling & vector databases
- Milvus Scalable open-source vector database for semantic search and RAG at large data volumes.
- Qdrant Fast vector database for semantic search and RAG back-ends.
- Label Studio Data labeling tool for text, images, audio and video, used to build and evaluate AI models.
Chroma FAQ
How much RAM does Chroma need?
The official minimum is 2 GB RAM. The docs say deploying Chroma with less than 2 GB of RAM is not recommended. The HNSW index must fit in RAM: for 1024-dimensional embeddings, max collection size in millions is about 0.245 x RAM in GB, plus at least 1 GB for the system. Disk should be at least as large as RAM plus several GB.
Does Chroma need a GPU?
CPU-only; the single-node sizing guide is based on CPU instances.
How much does it cost to self-host Chroma?
The software is free to download. Based on the official minimum 2 GB RAM: That fits a Hetzner Cloud CPX12 server at about €11.99/month (server only, excl. VAT). That covers the server only, not LLM API usage, backups, extra storage or your time.
Is Chroma free for commercial use?
Chroma is licensed under Apache-2.0 (Permissive). Internal business use is allowed under the license text. Permissive. Internal and commercial use allowed; keep license and notices. This is a summary, not legal advice.
What are common problems when self-hosting Chroma?
Chroma v1.0.0 and later no longer ship built-in authentication, and the server binds to 0.0.0.0 by default. Keep port 8000 on a private network or behind your own authenticating proxy. If a collection grows larger than available RAM, latency spikes as the OS swaps and the system quickly becomes unusable. Size RAM for the total number of embeddings you expect. Single-node Chroma does not scale forever; the docs suggest a distributed deployment when you exceed the single-node limits.
Is Chroma still maintained?
Yes, it is actively developed. Last commit: Oct 22, 2026. Latest release: 1.5.9 (May 5, 2026). Checked Oct 23, 2026.
How do I install Chroma?
The official method is Docker. The exact commands are in the "How to deploy" section above and in the official instructions: https://docs.trychroma.com/guides/deploy/docker
What are self-hosted alternatives to Chroma?
Other apps in Data labeling & vector databases: Milvus, Qdrant, Label Studio.
Sources
- GitHub repository: github.com/chroma-core/chroma
- Official documentation: docs.trychroma.com/
- License text: github.com/chroma-core/chroma/blob/main/LICENSE
- Hardware requirements, Gotcha 2, Gotcha 3: docs.trychroma.com/guides/performance/single-node
- Deploy instructions: docs.trychroma.com/guides/deploy/docker
- Gotcha 1: github.com/chroma-core/chroma/blob/main/docs/mintlify/reference/server-env-vars.mdx
- Hetzner Cloud pricing (cost estimate): www.hetzner.com/cloud/regular-performance
- GitHub stars, commits and releases: GitHub REST API, fetched
Facts on this page were checked against the project's own repository and docs on 2026-10-23. Spotted something out of date? Tell us.