Skip to main content

Documentation Index

Fetch the complete documentation index at: https://mintlify.com/AmiraliNotFound/dummy-gemini-bot/llms.txt

Use this file to discover all available pages before exploring further.

Dummy Gemini Bot is a production-grade, fully self-hostable Telegram bot powered by the Google GenAI SDK (Gemini). Built for developers, power users, and communities who want a deeply customizable AI companion, it combines multimodal chat (text, images, and voice notes), dual-engine Text-to-Speech, high-speed media downloading, and a React-based admin dashboard — all running inside a single Docker container. Out of the box, the bot ships with a witty, sarcastic Tehrani Persian persona, though every aspect of its personality and behavior is fully configurable through environment variables or the live admin web panel.
Ready to get the bot up and running right now? Jump straight to the Quickstart guide — you’ll have a live bot in under five minutes.

​
Key Features

AI Chat

Multimodal Gemini-powered conversations with async SQLite history, configurable context windows, dynamic model failover, and VIP persona overrides.

Text-to-Speech

Dual-engine TTS using Microsoft Edge TTS (free, natural Persian) and Google Gemini TTS (premium audio), with automatic failover between engines.

Media Downloader

Auto-detects Instagram, Threads, YouTube, and Pinterest links. Downloads and re-uploads media with inline streaming, album support, and a 50 MB bypass via Cobalt CDN.

Admin Dashboard

React-based Telegram Mini App with stats, moderation controls, broadcast tools, live settings editing, and real-time Gemini API quota meters.

Daily Summaries

Scheduled background task that auto-generates sarcastic Persian summaries of group chat activity and posts them at a configured time each day.

Support System

Built-in DM support ticket system. Users submit tickets via /support; admins respond with /reply. Includes voice note transcription via Gemini.

​
Project Structure

The repository is organized as a Python backend with a React frontend housed inside the webapp/ directory: 
dummy_gemini_bot/
├── src/
│   ├── __init__.py      # Package initializer
│   ├── config.py        # Settings loader, environment validations, and logging setup
│   ├── database.py      # Asynchronous database connection, schema setup, stats, and operations
│   ├── handlers.py      # Core Telegram handlers (Admin commands, Text/Multimodal messaging)
│   └── server.py        # Aiohttp Web API server, CORS, auth, and admin REST endpoints
├── webapp/
│   ├── src/
│   │   ├── App.jsx      # React WebApp frontend layout and dynamic configuration bindings
│   │   ├── index.css    # Design system, animations, and responsive styles
│   │   └── main.jsx     # Frontend entry point
│   ├── index.html       # WebApp template
│   └── vite.config.js   # Vite builder configurations
├── main.py              # Main entry point to initialize and start bot polling
├── requirements.txt     # Python dependency definitions
├── Dockerfile           # Multi-stage optimized Docker build definition (React + Python)
├── docker-compose.yml   # Compose setup for Docker volume and env mounts
├── .env.example         # Template configuration file
└── README.md            # Comprehensive system documentation

​
Requirements

​
Python Dependencies

The following packages are installed from requirements.txt:
PackagePurpose
python-telegram-bot>=20.0Async Telegram Bot API wrapper
google-genai>=0.1.1Google GenAI SDK (Gemini)
aiosqlite>=0.19.0Async SQLite database driver
python-dotenv>=1.0.0.env file loader
pillow>=10.0.0Image inspection and processing
aiohttp>=3.9.0Async HTTP client/server
aiohttp-cors>=0.7.0CORS middleware for the Web API
yt-dlp>=2024.04.09Video/media downloader
cryptography>=42.0.0Secure token and data operations
edge-tts>=6.1.9Microsoft Edge Text-to-Speech
curl-cffi>=0.10.0,<0.15.0Browser TLS impersonation for scraper bypass
instaloader>=4.10Instagram media scraping

​
System Dependencies

In addition to the Python packages, the following system-level tools are required:
  • ffmpeg — Used for audio format conversion (raw PCM → OGG), video thumbnail generation, and ffprobe metadata extraction. Installed automatically in Docker.
  • Deno runtime — Used by yt-dlp as its JavaScript engine for sites that require JS evaluation (e.g. certain YouTube flows). The Docker image copies the Deno binary directly from the official denoland/deno:bin image.

​
License

Dummy Gemini Bot is released under the MIT License. You are free to use, modify, and distribute it for personal or commercial projects. View the full source code and license on GitHub →

Build docs developers (and LLMs) love

Get started for freeTalk to us