Getting Started

elmtree chat is a chat interface for AI models with a strong focus on privacy. We do not analyze your chats. We do not store chats as server history: neither user messages nor LLM responses. We do not know your name. We send requests to LLM providers anonymously.

Main areas

• Chat is the workspace for messages and documents.
• Model selection keeps tasks on suitable strong models.
• Pseudonymization replaces detected personal data in inline documents and related follow-up questions before content is sent to a model.
• Web search helps with questions that need current information or sources.
• Account is required for regular use, admin features, and server-supported features.

Typical flow

1. Write a message or add a document.
2. For documents with sensitive data, check the pseudonymization preview.
3. Enable web search only when current information or sources are needed.

Local and hosted functions

Some functions run directly in the browser. Others, especially large documents, OCR, or RAG preparation, can optionally be supported by our server. The route depends on your settings and account.

More details are available under Privacy and Processing, Model Selection, and Pseudonymization.