parisneo/lollms-webui

★ 4,783PythonAudience · vibe coderComplexity · 3/5Setup · moderate

Mindmap

mindmap
  root((lollms-webui))
    What it does
      Local AI chat
      Image generation
      Music and video
      Personality routing
    Tech stack
      Python
      Docker and conda
      Ollama and vLLM
    Use cases
      Privacy-first AI
      Creative tools
      Expert assistants
    Audience
      Vibe coders
      Privacy-focused users

mindmap root((lollms-webui)) What it does Local AI chat Image generation Music and video Personality routing Tech stack Python Docker and conda Ollama and vLLM Use cases Privacy-first AI Creative tools Expert assistants Audience Vibe coders Privacy-focused users

Click or tap to explore — scroll the page freely

Things people build with this

USE CASE 1

Run AI language models privately on your own computer without sending data to external services

USE CASE 2

Choose from 500+ pre-configured AI personalities for tasks like coding help, medical symptom checking, or creative writing

USE CASE 3

Generate images using Stable Diffusion or DALL-E, music using MusicGen, or video using RunwayML directly from the chat interface

USE CASE 4

Route prompts to different AI models depending on task complexity to balance speed and quality

Tech stack

PythonpipDockercondaStable DiffusionOllamavLLMOpenAI API

Getting it running

Difficulty · moderate Time to first run · 30min

Requires cloning with Git submodules and downloading model files separately, Docker and conda setups are available as alternatives.

No license information was provided in the explanation.

In plain English

LoLLMs WebUI is a locally hosted web interface for running AI language models and other AI tools on your own computer, without sending your data to an external service. It is designed for a single user and supports connecting to a wide range of model sources, including locally stored model files, services like Ollama and vLLM, and external APIs from providers like OpenAI and Anthropic. You control which model handles each conversation. The interface provides a chat-style experience with a large library of pre-configured "personalities," which are sets of instructions that tune the AI toward specific roles. Examples described in the README include a medical assistant that helps identify symptoms, a legal advisor for interpreting contracts, a coding assistant, a creative story generator, and a music generation tool. According to the README, over 500 such expert configurations and more than 2,500 fine-tuned models are available through the system. Beyond text conversations, the tool supports generating images using backends like Stable Diffusion, DALL-E, and Midjourney, generating music using MusicGen, and generating video through services like RunwayML and CogVideoX. It also includes features for organizing discussions, searching through past chats, rating responses, and routing prompts to different models depending on task complexity. Installation is handled by a script for Windows, Linux, or macOS, or can be done manually using Python and pip. The repository has submodules, so cloning requires fetching those as well. The README notes that this project is in minimal maintenance mode and will eventually be replaced by a newer multi-user version of the project called lollms, which the same author maintains separately. Light and dark mode are available in the interface. Local discussion history is stored in a database on the user's own machine. Docker and conda setups are also supported.

Copy-paste prompts

Prompt 1

I want to set up lollms-webui on my Windows PC. Walk me through installing it using the installation script, connecting to an Ollama backend, and selecting a coding assistant personality.

Prompt 2

I am using lollms-webui and want to generate images using Stable Diffusion. How do I configure the image generation backend and trigger it from the chat interface?

Prompt 3

How do I create a custom AI personality in lollms-webui that acts as a Python tutor, giving step-by-step explanations for beginners?

Prompt 4

I want to route simple questions to a fast local model and complex ones to a larger model in lollms-webui. How does the task routing feature work?

Open on GitHub → Explain another repo

← parisneo on gitmyhub — every repo by this author, as a profile.

Verify against the repo before relying on details.