Skip to content

Home

Jump to bottom
Traliran edited this page Aug 13, 2026 · 4 revisions

🚀 Traliran AI Hub — Wiki

Welcome to the official documentation for Traliran AI Hub!

Traliran AI Hub is a lightweight, fully client-side web interface designed for seamless interaction with various Large Language Models (LLMs). The project focuses on speed, extreme privacy, and complete independence from heavy backends or build step tools.

🔑 Key Principles

  • KISS & Zero Build Step: Written in pure Vanilla JS, HTML, and CSS. No node_modules, Webpack, Vite, or heavy frontend frameworks. It runs instantly right in your browser.
  • Privacy First (100% Client-Side): Your API keys and conversation history are stored exclusively in your local browser storage (or your self-hosted database). API requests travel directly from your browser to the LLM providers—no middleman servers involved.
  • Multi-Provider Ecosystem: Full support for both self-hosted/local models (Ollama, llama.cpp) and cloud providers (OpenAI, Gemini, Claude, DeepSeek, Groq, OpenRouter, GLM, Qwen, and more).

📌 Navigation

  1. Quick Start & Hosting — Run instantly via GitHub Pages, local static servers, or self-hosted environments.
  2. Providers & API Setup — Step-by-step connection guides for every supported service:
    • Cloud Services: OpenAI, Anthropic Claude, Google Gemini, Groq, DeepSeek, OpenRouter, GLM (Zhipu AI), Qwen (DashScope).
    • Local Systems: Ollama, llama.cpp (configuring CORS, endpoints, and local options).
  3. PocketBase Integration — Setting up a lightweight PocketBase backend to synchronize chats, prompts, and settings across devices.
  4. Troubleshooting & FAQ — Fixing CORS blockages, Mixed Content errors ("Failed to fetch"), and key handling issues.
  5. 🛠️ Custom Builds, B2B Deployments & Integrations -

🔗 External Links

Clone this wiki locally