The ultimate all-in-one AI workspace, supercharged and fully portable.
Your fully private, customizable, and locally-hosted AI assistant. Chat with your documents, connect to any LLM, and manage your AI workspace without limits.
This is a customized, enhanced fork of the original AnythingLLM Desktop application, specifically optimized and packaged as a Portable Windows Application.
It transforms your desktop into a full-stack AI powerhouse, allowing you to interact with large language models (LLMs) and your own data securely. Because it's completely portable, you can run it from any folder or even a USB flash drive without touching your system registry or running complex installers.
While the original AnythingLLM is fantastic, this fork was created to provide a lightweight, no-install experience for Windows users, along with several under-the-hood enhancements.
(Note: Update these points with your actual specific changes)
- 🧳 True Portability: No installers, no hidden AppData folders. Everything runs straight from the extracted folder.
- ⚡ Optimized Performance: Faster document parsing and smoother UI rendering.
- 🎨 UI/UX Refinements: A cleaner, more intuitive interface tailored for power users.
- 🛠️ Expanded Customization: Deeper control over system prompts, chunking parameters, and RAG pipelines.
- 🏃♂️ Zero Installation: Download the archive, extract, and double-click
Main.exe. That's it. - 🔒 100% Privacy by Default: Run local models (via Ollama, LM Studio, etc.) and keep all your documents and chats strictly on your machine.
- 🧠 Bring Your Own LLM: Seamlessly switch between Local LLMs, OpenAI, Anthropic, Groq, Mistral, and more.
- 📚 Enterprise-Grade RAG: Drop in PDFs, Word docs, TXT files, or web links. The app automatically vectorizes them so you can "chat" with your data.
- 🗂️ Workspaces: Isolate context. Create separate workspaces for coding, creative writing, or financial analysis—each with its own documents and system prompts.
Since this is a portable version, getting started takes less than a minute.
- Go to the Releases page.
- Download the latest
.ziparchive. - Extract the
.zipfile to any folder on your computer (or a portable drive). - Double-click
Main.exeto launch the application. - Enjoy your local AI workspace!
Language Models (LLMs)
- Local: Ollama, LM Studio, LocalAI, Llama.cpp
- Cloud: OpenAI, Anthropic (Claude), Groq, Google Gemini, Mistral, TogetherAI, OpenRouter
Vector Databases
- LanceDB (Default, built-in)
- Pinecone
- Chroma
- Qdrant
- Milvus
Embedding Models
- Built-in Native Embedder (No API needed)
- Ollama
- OpenAI
- Cohere