chroma

Store embeddings and metadata for vector and full-text search.

Updated Apr 19, 2026

Search Tags:#rag #embedding #semantic-search #vector-database #self-hosted #document-retrieval #metadata-filtering

Ready to run

npx skills add https://github.com/gqf2008/hermez-ai --skill chroma-gqf2008

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: chroma
Source: https://github.com/gqf2008/hermez-ai/tree/main/skills/mlops/chroma
Command: npx skills add https://github.com/gqf2008/hermez-ai --skill chroma-gqf2008

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

Open-source, self-hosted embedding database for AI applications to store embeddings and metadata, enabling efficient vector and full-text search with metadata-based filtering.

Core Features & Use Cases

Store embeddings and associated metadata for AI tasks.
Perform vector and full-text search with fast retrieval and filters.
Use cases include semantic search, RAG workflows, and document retrieval in local/development setups.
Lightweight, scalable API suitable for notebooks to production clusters.

Quick Start

Install chromadb and create a collection to run a basic similarity search.

Frequently Asked Questions about chroma

High-intent search queries and answers about installing and using this skill.

FAQPage Schema

How do I store embeddings and metadata for a local RAG workflow?▼

You can store embeddings and metadata for local RAG workflows using a self-hosted vector database that manages document retrieval and metadata filtering efficiently.

What is the best way to run semantic search on my own documents?▼

The best way to run semantic search on your own documents is using a self-hosted embedding database that supports vector search and metadata filtering across local setups.

Can I use a vector database for document retrieval in production clusters?▼

Yes, you can use this lightweight, scalable vector database for document retrieval in production clusters, as it transitions smoothly from notebooks to production environments.

Does a self-hosted embedding database support full-text search?▼

A self-hosted embedding database can support both fast vector and full-text search, allowing you to retrieve relevant documents using multiple search methods.

How does metadata filtering work in vector search?▼

Metadata filtering in vector search works by storing associated metadata alongside embeddings, allowing you to apply specific filters to narrow down your document retrieval results.

Do I need a complex API to manage a local vector store?▼

No, you do not need a complex API to manage a local vector store, as this solution exposes a simple four-function API to handle embeddings and search operations.