Serge

Serge is an open-source platform for running large language models (LLMs) locally or on your own infrastructure, giving teams full control over their data and AI workflows. It provides a web-based interface for chatting with and managing models, enabling organizations to deploy private, self-hosted AI assistants without relying on third-party cloud services. Built for technical and privacy-conscious users, Serge focuses on reproducible, transparent, and configurable LLM operations.

The platform supports popular open-source models via backends like llama.cpp, allowing you to download, load, and switch between models directly from the interface. Serge includes tools for model configuration (such as temperature, context length, and system prompts), conversation history management, and multi-user access. It exposes APIs for integration with existing systems and workflows, enabling developers to embed LLM capabilities into internal tools, scripts, or applications. Because it is self-hosted, Serge can be deployed on-premises, in private clouds, or on developer machines, making it suitable for environments with strict compliance or data residency requirements.