Introduction
Welcome to Mixedbread
Mixedbread transforms how you work with data by making everything searchable through AI. Whether it's PDFs buried in folders, product catalogs, code repositories, or support documentation, our multimodal AI models understand and index any content type. Create intelligent search experiences that understand natural language queries and find exactly what users or AI agents need, regardless of how they ask.
Quickstart Guides
Get up and running in under five minutes with these practical guides:
Vector Stores
Transform any data into an AI-native search engine that understands natural language queries and powers AI agents.
Embeddings
Convert any content into vector representations with our embedding models.
Reranking
Improve search relevance by reordering results based on semantic understanding.
Document Parsing
Extract LLM-ready content from documents with our intelligent parsing tools.
Learn the Fundamentals
Understand the theory and concepts powering modern search and retrieval systems:
Core Concepts
Understand how AI-native search works and why it's transforming data discovery.
Engineering Blog
Explore tutorials, benchmarks, and best practices from our engineering team.
Explore Our Models
Leverage our state-of-the-art open-source model families:
Embedding Models
Explore our multimodal models that understand text, images, and documents across languages.
Reranking Models
Explore our powerful v2 rerankers that significantly boost search quality.
Build with Our API & SDKs
Integrate Mixedbread into your applications with our developer tools:
API Reference
Comprehensive documentation of endpoints, parameters, response formats, and error codes.
Client SDKs
Official Python and TypeScript libraries for seamless integration.
Join Our Community
Discord
Get support, share your projects, and connect with other developers.
Stay updated with the latest features, models, and community highlights.
Follow our company updates and technical content for professionals.
Last updated: June 11, 2025