zsj1029

zsj1029 / LiteChat

Public

LiteChat is forked from llama.cpp WebUI, Make It supports vLLM

18
2
100% credibility
Found May 06, 2026 at 18 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
TypeScript
AI Summary

LiteChat is a web-based chat interface for local AI models, supporting file uploads, multi-model selection, real-time streaming, conversation branching, and integration with external tools via MCP.

How It Works

1
🖥️ Open the chat window

Launch the simple web chat app in your browser to start talking to AI.

2
🤖 Pick your AI brain

Choose from available AI personalities so it understands your needs best.

3
💬 Send your first message

Type a question, add pictures or files, and watch the AI think and reply right away.

4
🛠️ Add smart helpers

Use ready prompts or tools from connected services to tackle bigger tasks.

5
🔄 Refine the conversation

Edit messages, try again, or explore different paths to get perfect answers.

✨ Your AI companion shines

Enjoy private, helpful chats that feel natural and get things done.

Sign up to see the full architecture

4 more

Sign Up Free

Star Growth

See how this repo grew from 18 to 18 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is LiteChat?

LiteChat is a TypeScript-based web UI forked from llama.cpp's WebUI, delivering a clean browser interface for chatting with local LLMs. It connects to llama-server or vLLM backends, handling multimodal inputs like images, PDFs, audio, and text files with previews and drag-drop uploads. Developers get instant chats with model switching, without wrestling Electron apps or clunky terminals.

Why is it gaining traction?

It stands out with agentic workflows, tool calls, and MCP prompt/resource integration via `/` and `@` shortcuts, plus conversation branching for edits/regenerates. File handling shines—PDFs render as text or images, audio records directly—making multimodal testing seamless on vision/audio models. The hash-router static build deploys anywhere, beating heavier UIs in speed and portability.

Who should use this?

Local LLM hobbyists experimenting with vLLM or llama.cpp on consumer hardware. Devs prototyping agentic apps needing quick file/vision support. Teams avoiding SaaS for privacy-focused RAG chats with attachments.

Verdict

Grab it for lightweight local AI chats—features punch above its 18 stars—but the 1.0% credibility score signals early maturity; sparse docs and unproven tests suit tinkerers, not prod. Fork and contribute if it clicks.

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.