ExpressTokyo v2.0
Stealth AI Assistant — powered by Groq (llama-3.3-70b + llama-4-scout vision) via secure Vercel proxy. No local API keys needed.
Install
npm install -g expresstokyo
Run
expresstokyo
Auto-installs Python 3.12 and all dependencies if they're missing.
Hotkeys
Ctrl+C | Copy text → AI answers instantly |
Alt+S | Screenshot → AI solves the screen |
Alt+T | Snipping tool → OCR / extract text |
Alt+R | Retry last question |
Alt+V | Auto-type clipboard answer |
Alt+M | Mash mode (keypress = next char) |
Alt+X | Toggle code preview window |
Alt+↑/↓ | Scroll code preview |
` | Show overlay / Toggle visibility |
Esc | Hide overlay |
Ctrl+Q | Quit completely |
Modes
Alt+1 | Aptitude |
Alt+2 | DSA / Coding (default) |
Alt+3 | Fullstack |
Alt+4 | AWS |
Alt+5 | Copy-Paste (raw OCR) |
Requirements
- Windows (tested on Win10/11)
- Node.js ≥ 14
- Internet connection
Python 3.12 + all Python deps (pyperclip, keyboard, requests, pillow, pyautogui) are auto-installed on first run.
Powered by
- Groq — ultra-fast LLM inference
llama-3.3-70b-versatile for text
meta-llama/llama-4-scout-17b-16e-instruct for vision/screenshots