Skip to content
Triplemblem

Project

LLM Framework

A self-hosted framework for running language models locally on your own device. It organizes one model into separate workspaces, each with its own prompts, documents, memory, and permissions.

LLM Framework is an open-source, self-hosted workspace for running language models locally on your own device without tying you to one model, profession, or fixed collection of assistants.

Framework goal

One shared model can be organized into separate workspaces. Each workspace can use its own instructions, retrieved documents, memory, conversations, tools, and permissions without mixing unrelated context.

The current framework runs with Docker, Ollama, and PostgreSQL. It is under active development and is not yet hardened for multi-user or public-internet deployment.