AnythingLLM
FreemiumThe open-source, all-in-one AI application for private document chat, RAG, and AI agents with any LLM.
AnythingLLM is an open-source, all-in-one AI application built by Mintplex Labs that lets individuals and teams chat with their documents, run AI agents, and connect virtually any LLM without sending data to third parties. It packages document ingestion, vector search, model routing, agent skills, and a developer API into a single installable desktop app or Docker deployment. A managed cloud tier is also available for teams that prefer a hosted private instance.
What is it
A privacy-first AI workspace that combines retrieval-augmented generation, AI agents, and multi-LLM support into one local or self-hosted application.
What it can do
Ingest PDFs, Word files, CSVs, codebases, and web pages; answer questions over them with RAG or full-text attachment; build no-code agents with web search, scraping, SQL, and file-system skills; route different prompts to different models; and expose everything through a REST API or embeddable widget.
Who is it for
Solo knowledge workers, researchers, developers, and small-to-medium teams who want ChatGPT-like capabilities over private data while keeping control of their documents, models, and infrastructure.
Key Features
Document Ingestion & RAG — Chat with Any Document
Upload PDFs, DOCX, TXT, CSV, code repositories, and web pages into isolated workspaces. AnythingLLM automatically chunks, embeds, and indexes them using a built-in vector database, then retrieves the most relevant passages to ground answers in your actual content.
Flexible Document Modes — Embed, Attach, or Pin
Choose embedding for repeated semantic retrieval across large documents, full-text attachment for one-off precise questions that fit the context window, or document pinning to force critical files into every prompt. Each mode trades speed, cost, and accuracy differently.
No-Code AI Agents — Built-In Skills and Agent Flows
Activate agents with `@agent` or let the app detect tool-capable models automatically. Agents can search embedded documents, browse the web, scrape pages, save files, summarize documents, generate charts, query SQL databases, and interact with the file system without writing code.
Dynamic Model Router — Route Prompts to the Right LLM
Define rules that send different kinds of questions to different providers or models. Route math to a reasoning model, translations to a fast multilingual model, and legal questions to your most capable model, saving money while matching capability to the task.
Use Cases
Internal knowledge bases for teams
Upload company policies, procedures, and documentation so employees can ask natural-language questions instead of searching shared drives.
Private research libraries for academics
Embed hundreds of papers and instantly surface relevant methodology, findings, or citations across a collection without manual keyword searches.
Local document Q&A for sensitive industries
Healthcare, finance, and legal teams keep contracts, records, and reports on their own servers while still enabling AI-powered question answering.
Developer testing across LLM providers
Compare how different local and cloud models answer the same codebase or documentation corpus before choosing a provider for production.
Pricing plans
Frequently Asked Questions
Discussion
No comments yet. Be the first to start the thread.