Chroma

What Chroma does

Chroma is an open-source vector database company that provides search infrastructure for AI applications. Its core product, ChromaDB, is an Apache 2.0-licensed database that unifies dense vector search, sparse vector search (BM25, SPLADE), full-text search, regex matching, and metadata filtering behind a single API. The system is built on object storage (S3/GCS) with query-aware automatic data tiering across hot (memory), warm (SSD), and cold (object) layers, which the company claims delivers up to 10x cost reduction versus memory-resident legacy search systems. Chroma offers Python, JavaScript/TypeScript, and Rust SDKs and integrates with LangChain, OpenAI, Cohere, HuggingFace, and a Model Context Protocol server for AI coding agents.

Chroma operates a dual-distribution model. The open-source database drives adoption through developer communities (27k GitHub stars, 90k+ dependent codebases, 10k+ Discord members, 15M+ monthly downloads). Monetization comes from Chroma Cloud, a managed serverless service that reached general availability in August 2025, with a $5 free credit tier, Pro plans, and an Enterprise tier offering BYOC VPC deployment, multi-region replication, customer-managed encryption keys, AWS PrivateLink, and custom SLAs. Chroma Cloud holds SOC 2 Type II certification. Named enterprise customers include Capital One and UnitedHealthcare; startup customers include Mintlify, Weights & Biases, Conduit, Propel, Medwise, and Cofounder.

In 2025-2026, Chroma substantially expanded its product surface, launching Sparse Vector Search and Chroma Sync (GitHub ingestion) in October 2025, Web Sync in November 2025, CMEK in December 2025, and Package Search MCP in September 2025, which lets AI coding agents search indexed source code across NPM, PyPI, Go, Crates.io, RubyGems, and Terraform. In March 2026, the company released Context-1, a 20-billion parameter open-source Mixture of Experts agentic search model co-developed with UIUC and UC Berkeley, achieving 10x faster inference and 25x lower cost than frontier models at comparable accuracy. The company was founded in 2022, is headquartered in San Francisco, and has raised approximately $20.3M in total disclosed funding across a 2022 pre-seed and an $18M seed round in April 2023 led by Quiet Capital.

Name

Chroma

Legal name

Chroma Inc.

Website

https://trychroma.com

Company type

Private

Founded year

2022

Operating status

Operating

Headcount range

1–10 employees

Short description

Chroma is an open-source vector database company providing search infrastructure for AI applications, combining dense, sparse, full-text, and metadata search on object storage. It serves developers and enterprises through its Apache 2.0 ChromaDB, managed Chroma Cloud service, and BYOC Enterprise tier, with named customers including Capital One and UnitedHealthcare.

Ownership category

akta.pro rank

Product category

Vector Database / AI Search Infrastructure

NAICS

Software Publishers (5132)

SIC

Services-Prepackaged Software (7372)

akta.pro primary industry

Enterprise Search, Indexing & Content Discovery (HDAEAGAH)

akta.pro secondary industry

Database/Application Data Layer Modernization (DB migration, ORM, caching) (BPAEAFAL)

Keywords

• Vector database
• AI search infrastructure
• Embedding database
• Retrieval-augmented generation
• Semantic search platform

Headquarters

HQ city

San Francisco

HQ country

United States

HQ region

North America

Offices1 record

Markets served

GTM type

B2B

Offering type

Software

Cost components

Technology or R&D, Personnel, Infrastructure, Marketing or Sales, Operations

Revenue model

1. Chroma Cloud (Managed Service): Serverless managed vector database service with automatic scaling, zero-ops infrastructure, and object storage-based pricing. Offers free tier with $5 in credits for new users. Paid plans include Pro with direct Slack support and Enterprise with custom SLAs, BYOC VPC deployment, multi-region replication, and 24/7 assistance.
2. Open Source (Self-Hosted): Apache 2.0 licensed open-source database available for local deployment and self-hosting. Revenue generated through community adoption leading to enterprise conversions and cloud service adoption.

Pricing tiers

Go-to-market motion2 records

Distribution channels4 records

Marketing channels8 records

Core offering

Chroma provides open-source vector database infrastructure for AI applications, supporting vector similarity search, sparse vector search (BM25, SPLADE), full-text search, regex pattern matching, and metadata filtering in a single unified API. The platform is available as an Apache 2.0 open-source database, a managed serverless cloud service, and an enterprise BYOC (Bring Your Own Cloud) deployment. The offering is extended by Context-1, a 20B parameter open-source agentic search model, and Package Search MCP, an integration for AI coding agents.

Product overview

Chroma is an open-source search infrastructure company offering a unified AI data platform built around its core ChromaDB vector database. The portfolio consists of Chroma Database (the open-source Apache 2.0 vector database), Chroma Cloud (managed serverless service), Chroma Enterprise (BYOC VPC deployment), Chroma Sync (data ingestion from S3/GitHub/Web), and Context-1 (a 20B parameter open-source search model). ChromaDB supports vector similarity search, full-text search (BM25/SPLADE), regex, and metadata filtering, with performance reaching ~20ms p50 latency at 100K vectors and scaling to billions of vectors. Package Search MCP extends the platform to AI coding agents for code search. All offerings are built on object storage with automatic data tiering, targeting developers building RAG applications, AI agents, and semantic search systems.

Differentiator

Problem solved

Functional benefit

Brands

• Chroma Cloud: Managed, serverless vector database service with automatic scaling and data tiering.
• Context-1
• Package Search MCP
• Chroma Sync

Products and services

• Chroma Database (ChromaDB) Open-source vector database providing fast, serverless, and scalable infrastructure supporting vector, full-text (BM25, SPLADE), regex, and metadata search. Built on object storage with automatic data tiering, delivering ~20ms p50 query latency at 100K vectors (384 dim) and scaling to billions of vectors with 90-100% recall. Available as open source (Apache 2.0) or self-hosted.
• Chroma Cloud Fully managed, serverless vector database service providing zero-ops deployment with automatic scaling. Features SOC 2 Type II compliance, private networking (AWS PrivateLink), customer-managed encryption keys (CMEK), and serverless usage-based pricing. Hosts over 1.2M collections.
• Chroma Enterprise BYOC (Bring Your Own Cloud) deployment option enabling enterprises to run Chroma within their own VPC while Chroma manages the control plane. Offers multi-cloud/multi-region replication, point-in-time recovery, custom SLAs, and 24/7 support with dedicated Slack communication.
• Chroma Sync Data synchronization service for Chroma Cloud enabling automatic ingestion from external sources including S3 buckets, GitHub repositories, and web pages. Automatically chunks, embeds, and indexes data without manual pipeline configuration.
• Package Search MCP Model Context Protocol (MCP) server enabling AI coding agents to perform semantic and regex-based search across thousands of indexed open-source package dependencies. Provides tools for semantic search, regex pattern matching, and file-level reading. Indexes ~13k versions of 3k packages, with embeddings powered by Modal.
• Context-1 20-billion parameter open-source AI search model optimized for retrieval-augmented generation and agentic search. Uses Mixture of Experts architecture with a Self-Editing Context mechanism achieving 0.94 accuracy in document pruning. Delivers comparable accuracy to frontier models (73% on Harness-1 benchmark vs GPT-5.4's 70.9%) at 10x faster inference and 25x lower operational costs.

Quantifiable outcome

• 10x faster inference than frontier models for retrieval tasks
• +6 more outcomes

Named customers7 records

Segments4 records

Ideal customer profiles3 records

Technology focussed Yes

API detail

Has API

Yes

API docs

API detail

Core technology

AI maturity

App detail

Integration1 record

AI capability7 records

Feature8 records

Partnerships

Eight partnerships are on record, tiered core and minor.

• Amazon Web Services (AWS)coreTechnology or IntegrationPrimary cloud infrastructure provider for Chroma Cloud. Chroma is hosted on AWS across multiple Availability Zones for redundancy. AWS provides compute, storage, and networking infrastructure under a shared responsibility model for security.
• Google Cloud Platform (GCP)coreTechnology or IntegrationSecondary cloud infrastructure option with EU data processing location available for customers requiring EU data residency.
• OpenAIcoreTechnology or IntegrationAI services partner providing embedding model access. Used when customers perform semantic queries and select OpenAI embedding models.
• Google GeminicoreTechnology or IntegrationAI services partner for embedding and semantic query capabilities. Available when customers specifically select Gemini providers in the UI.
• StripecoreTechnology or IntegrationPayment processing provider for subscription billing and payment collection on Chroma Cloud paid plans.
• ModalcoreTechnology or IntegrationCompute services provider powering Package Search embeddings. Modal provides the underlying compute for indexing and serving package search results.
• University of Illinois at Urbana-ChampaignminorStrategic or Co-development PartnerResearch collaboration on Harness-1, a 20-billion parameter open-source AI search agent. Chroma co-developed the model with UIUC and UC Berkeley researchers, achieving 73% average accuracy vs GPT-5.4's 70.9%. Model released under Apache 2.0 license.
• UC BerkeleyminorStrategic or Co-development PartnerResearch collaboration on Harness-1 open-source AI search agent alongside UIUC and Chroma. Contributed to state-externalizing harness architecture enabling frontier-level performance at reduced computational cost.

Scale indicators13 records

Recent moves6 records

Expansion highlights6 records

Direct peers

• Pinecone: Pinecone is a leading managed vector database directly competing with Chroma for the same RAG and semantic-search workloads. Both target enterprise AI developers and offer serverless managed services with hybrid search capabilities, differing mainly on architecture (Pinecone uses proprietary pods vs Chroma's object-storage tiering).
• Qdrant: Qdrant is an open-source vector database written in Rust that competes head-on with Chroma for production-grade vector search workloads. Both offer hybrid search, filtering, and managed cloud offerings targeted at AI engineers building RAG and semantic search applications.
• Weaviate: Weaviate is an open-source vector database (BSD-licensed) competing directly with ChromaDB on developer mindshare, hybrid search, and enterprise deployments. Both companies were part of the same 2023 vector database funding surge and target overlapping AI engineer and enterprise personas.
• Milvus: Milvus is a popular open-source vector database (LF AI & Data graduate) directly comparable to ChromaDB on scalability, hybrid search, and cloud-managed deployments. Both target billion-scale vector workloads and have substantial open-source community traction among AI developers.

Broad incumbents

• Vespa: Vespa is Yahoo's open-source big-data serving engine that combines vector search, lexical search, and structured query at scale. It overlaps with Chroma's unified search positioning but is a broader serving platform aimed at large-scale recommendation and search workloads.
• Elasticsearch: Elasticsearch is a broad-incumbent search and analytics engine that has added dense_vector and semantic search capabilities directly competing with Chroma's hybrid search. Many enterprises already run Elasticsearch, creating a strong incumbency advantage for hybrid search workloads.
• MongoDB Atlas Vector Search: MongoDB Atlas Vector Search adds vector search capabilities inside the broader MongoDB document database platform. It is a broad incumbent threat because existing MongoDB customers can adopt vector search without introducing a separate vendor like Chroma.
• AWS OpenSearch: AWS OpenSearch is the hyperscaler's managed search and analytics service with k-NN vector capabilities, positioning it as a broad incumbent competitor. Because Chroma itself runs on AWS, OpenSearch represents both a partner-coexistence opportunity and a competitive threat via bundled enterprise pricing.
• Databricks Mosaic AI Vector Search: Databricks offers native vector search integrated with its data lakehouse and ML platform, competing with Chroma for enterprise RAG and AI workloads. It leverages an incumbent data platform presence to bundle vector capabilities alongside enterprise data, governance, and ML tooling.

Emerging players

• LanceDB: LanceDB is an embedded, open-source vector database built on the Lance columnar format, targeting AI developers who need serverless, in-process vector search. It overlaps with Chroma's developer-focused RAG use cases but differentiates through a lightweight embedded deployment model and columnar storage.

Market position

Strengths5 records

Weaknesses5 records

Competitive moat6 records

Key risks6 records

Key highlights7 records

Customer concentration

• Xx.com/trychroma
• GitHubgithub.com/chroma-core/chroma
• YouTubeyoutube.com/@trychroma

Compliance1 record

Revenue estimate

Valuation estimate

Number of profiles

Profiles2 records

Funding overview

Funding rounds2 records

Investors2 records

Funding detail is available on the Subscription and Enterprise plan.Contact sales →

M&A

Investments

M&A and investment is available on the Subscription and Enterprise plan.Contact sales →

Frequently asked questions about Chroma