Chroma

Chroma is a company developing an open-source embedding database (or vector database), also known as Chroma. The database aims to simplify the development of large language model (LLM) apps by making knowledge, facts, and skills pluggable for LLMs. Chroma offers tools to store embeddings and their metadata, embed documents and queries, and search embeddings. Chroma consists of a Python client SDK, JavaScript/TypeScript client SDK and a server application. In Python, Chroma can run in memory or in client/server mode. In JavaScript, Chroma runs in client/server mode and talks to a Python back end.

Diagram showing Chroma interacting with an LLM app.

The Chroma project is coordinated by a small team of full-time employees based in Potrero Hill. The company was cofounded by Jeff Huber and Anton Troynikov in April 2022. The Chroma database is released under an Apache 2.0 license. Chroma has stated any feature an individual developer would find useful will always be open source, an approach called buyer-based open source. The company plans to provide a commercial product in the future with exclusive functionality.

Chroma integrates with other frameworks in the AI stack, including LangChain and LlamaIndex (formerly GPTIndex). LangChain provides a framework for prototyping LLM applications locally, and Chroma provides a vector store to simplify running apps locally. The collaboration between LangChain and Chroma was announced on February 13, 2023. Users can import Chroma when using LangChain. LlamaIndex has a dedicated Chroma Loader that returns a set of texts corresponding to embeddings retrieved from a Chroma Index. The user initializes the loader with a Chroma index and then passes in a query vector. The loader is designed as a way to load data into LlamaIndex and/or subsequently used as a tool in a LangChain Agent.

Other integrations include:

• ScienceIO—an A.I-native developer toolkit for healthcare
• Replit—a browser-based IDE with AI functionality

On April 7, 2023, Chroma announced an $18M seed round led by Astasia Myers from Quiet Capital with participation from angels investors that include Naval Ravikant, Max and Jack Altman, Jordan Tigani (Motherduck), Guillermo Rauch (Vercel), Akshay Kothari (Notion), Amjad Masad (Replit), Spencer Kimball (CockroachDB), and other founders and leaders from ScienceIO, Gumroad, MongoDB, Scale, Hugging Face, and Jasper.

As part of the seed round announcement, the company also announced a previous pre-seed round from May 2022 led by Anthony Goldbloom (Kaggle) from AIX Ventures, James Cham from Bloomberg Beta, and Nat Friedman and Daniel Gross (AI Grant).