AnythingLLM releases v1.18.0
Fixes embed-origin allowlists, required tool parameters and attachment-state isolation.
AnythingLLM ↗Product profile →TOOLORA / Local reading edition
AnythingLLM
Chat, cited document retrieval and configurable agents around local models and workspaces.
Overview
AnythingLLM combines model chat, document ingestion and cited knowledge retrieval in workspaces, with agent tools and configurable flows. The desktop app targets personal use, while Docker self-hosting adds multi-user and permission features; their feature sets differ. Use local models or explicitly configure cloud models, vector stores and external services. It supplies an application interface rather than requiring a document pipeline to be assembled from scratch.
- Company country / region
- United States · Registered office
Usage guide & costs
- Before you start
Core code is MIT. Self-hosted data stays in your environment by default, but configured external models, databases or tools transmit data. Telemetry can be disabled. Prepare models/dependencies for offline use, and distinguish desktop, cloud and mobile terms.
Official details ↗ · Official details ↗ · Official details ↗ · Official details ↗
Update history
Official sources and references
- Official websitehttps://anythingllm.com/download
- Documentationhttps://github.com/Mintplex-Labs/anything-llm/blob/master/README.md
- Documentationhttps://github.com/Mintplex-Labs/anything-llm/blob/master/TERMS_SELF_HOSTED.md
Descriptions summarize public information; consult official websites for features, pricing and regional availability.