My AI Studio

Full-stack AI chat workspace — multi-provider LLMs, streaming, sessions, and a 9-language UI.

Python FastAPI React Vite Tailwind Radix UI

Demo

Model connection flow and streaming chat (from the open-source repo):

My AI Studio — model connection and setup

My AI Studio — streaming chat session

What it is

A production-style AI chat workspace you can run locally or deploy: FastAPI backend for auth, sessions, and streaming chat; React frontend with Radix UI and Tailwind.

Highlights

Bring-your-own-key — DeepSeek, OpenAI, Gemini, Qwen, OpenRouter, or Ollama; model ID and API key in a connection dialog
Streaming SSE — token-level chat with session history and system prompts
9-language i18n — EN, 简/繁中文, 日本語, 한국어, Español, Français, Deutsch, Русский; locale from IP/browser with user override in localStorage
Security-minded — no keys in repo; encrypted per-user configs; see SECURITY.md in the repo

Why it’s useful

Portfolio-grade full-stack AI UI: auth, persistence, provider abstraction, and internationalization in one repo—not a single-file script.

Agent loop concepts →