Lollms WebUI

A local LLM web interface with GGUF, Ollama and cloud API support, local chat storage, and image, video and music generation.

Lollms WebUI is a local AI interface for writing, coding assistance and media generation through a choice of models. The project receives minimal maintenance, and its interface has no built-in user authentication.

Local and hosted models You can use Hugging Face models, GGUF/GGML and ExLlamaV2, connect to Ollama or vLLM, or send requests to OpenAI, Anthropic and OpenRouter.
Task-specific chat Personalities provide preset behavior for different tasks, while Smart Routing chooses models according to prompt complexity, response speed or service cost.
Image and video work Image analysis and generation sit alongside video generation, with integrations including Stable Diffusion, Flux, ComfyUI and CogVideoX; hosted choices include DALL-E, Midjourney and RunwayML.
Music generation MusicGen provides a separate option for creating music.
Discussion management Chats live in a local database, with search and export, message editing and answer ratings.

Available generation capabilities depend on the models and services you connect. The interface runs locally, but requests using cloud services go to those providers. Docker deployment is supported, and headless mode exposes the generation API without the full web interface.

Lollms WebUI is free and open source under Apache 2.0. It's intended for local use; the separate lollms project provides multi-user and MCP capabilities.