Repolode
Submit
chroma-core

chroma-core/chroma

Open-source embedding database for AI apps, storing documents and vectors to power semantic, hybrid and full-text search.

View on GitHubWebsite
Stars
29,535
Forks
2,571
Open issues
951
Last commit
7 hours ago

What is it?

Chroma is an open-source data infrastructure layer for AI applications. It handles tokenization, embedding and indexing of documents added to a collection, so developers can store text and query it by meaning rather than exact keywords.

The core API centers on a small set of collection operations: create, add, query and get. Documents can carry metadata used for filtering, and queries can also filter on document contents. Chroma can run in-memory for prototyping, with persistence available, or in client-server mode started from the command line. Official clients are provided for Python and JavaScript.

The project also offers Chroma Cloud, a hosted service for serverless vector, hybrid and full-text search. The core codebase is written primarily in Rust and released under the Apache-2.0 license.

Key features

  • 01Automatic tokenization, embedding and indexing of added documents
  • 02Simple collection API: create, add, query and get
  • 03Metadata and document-content filters on queries
  • 04In-memory mode for prototyping, with optional persistence
  • 05Client-server mode and official Python and JavaScript clients

Use cases

  • Building retrieval-augmented generation (RAG) pipelines over private documents
  • Adding semantic search to an application's text content
  • Storing embeddings for AI agents that need long-term memory lookup

Quick start

  1. 1

    Install the Python client

    pip install chromadb
  2. 2

    Start client-server mode (optional)

    chroma run --path /chroma_db_path

Things to know

  • The README does not cover production hosting details; self-hosting requires setup effort.
  • Chroma Cloud is a paid hosted service beyond the free credits.

Topics

#vector-database#embeddings#semantic-search#rag#llm#rust#agents#ai#ai-agents#database

FAQ

What is chroma?
Open-source embedding database for AI apps, storing documents and vectors to power semantic, hybrid and full-text search. Chroma is an open-source data infrastructure layer for AI applications. It handles tokenization, embedding and indexing of documents added to a collection, so developers can store text and query it by meaning rather than exact keywords.
Is chroma free to use? What license does it have?
chroma is open source under the Apache-2.0 license. Check the license terms before using it in commercial products.
Is chroma still maintained?
Actively maintained. The last commit to chroma was on Oct 20, 2026.
What can I use chroma for?
Building retrieval-augmented generation (RAG) pipelines over private documents Adding semantic search to an application's text content Storing embeddings for AI agents that need long-term memory lookup
How do I install chroma?
Install the Python client: pip install chromadb Start client-server mode (optional): chroma run --path /chroma_db_path
Where can I get chroma?
The source code and installation instructions are on GitHub: https://github.com/chroma-core/chroma

Summary generated by AI from the project's README. GitHub stats refresh daily.

Related projects