Gemi Chat Bot

Gemi is a Telegram chat bot that leverages the gemini-pro API for text-based message generation and Bing for image generation. It is designed to interact with users by processing text and image inputs and responding appropriately.

Features

• Text conversation with users using the gemini-pro generative model.
• Image generation based on prompts using Bing.
• Search query handling with informative responses.
• PDF and document understanding capabilities.
• Built-in logging and error tracking via Sentry.
• Extendable with middleware and custom route handlers.

One-Click Deployment

Use the button below to deploy your own Gemi bot on Render.

Note: You need a Render account to use this. Or, create your free account here.

Quick Start

To get started with Gemi chat bot, follow these steps:

Prerequisites

• Python 3.10+
• A Telegram bot token from BotFather
• API keys for gemini-pro from here
• Auth cookies of Bing AI from here, used for image generation (optional)
• A configured webhook host. For localhost, follow this. (optional)

Installation

1. Clone the repository:

git clone https://github.com/Akash98Sky/gemi-chat-bot.git
cd gemi-chat-bot

2. Install the required dependencies:

pip install -r requirements.txt

Configuration

Create a .env file at the root of the project directory with the following environment variables:

BOT_TOKEN=<your_telegram_bot_token>
GOOGLE_API_KEY=<your_gemini_pro_api_key>
BING_COOKIE=<your_bing_auth_cookie>

# Webhook settings to wake up the bot (optional) - required only if your service spins down while idle (e.g.: Heroku, Render)
APP_HOSTNAME=<your_webhook_host> (e.g.: abc.xyz.com)
WEBHOOK_SECRET=<your_webhook_secret>

# Voice API settings
VOICE_API_URL=<your_voice_api_url>
TTS_VOICE=festival:cmu_us_slt_arctic_hts
STT_ENGINE=vosk

Running the bot

Use gunicorn to launch the application:

gunicorn main:web_app --bind 0.0.0.0:8080 -k aiohttp.GunicornWebWorker

Replace 8080 with the port you want to run your web server on.

Running the Gemi voice API

If you want to use the Gemi voice API, follow the instructions below:

• Clone the Gemi voice repository.

  git clone https://github.com/Akash98Sky/gemi-voice.git

• Install all its dependencies.

  cd gemi-voice && pip install -r requirements.txt

• Run the voice API on a different port.

  uvicorn main:app --host 0.0.0.0 --port 2024

• Update the gemi-bot .env file with the voice API URL.

  VOICE_API_URL=http://0.0.0.0:2024
  TTS_VOICE=edge:en-IN-NeerjaExpressiveNeural
  STT_ENGINE=vosk

Webhook on localhost

• Have ngrok installed on your local machine. If you don't have it installed, download it from ngrok's website and follow their installation instructions.

• Ensure your local server is ready to receive webhooks.

• Steps

  1. Start Your Local Server Ensure that your local server is running on a specific port. For example, if you're using a Node.js Express server, it might be running on port 8080.

  2. Run Ngrok Open a new terminal window and run ngrok to expose your local server to the internet. Replace 8080 with the port number your local server is using.

  ngrok http 8080

  3. Copy the Ngrok URL Once ngrok is running, it will display a public URL (e.g., https://abc123.ngrok.io). Copy this URL; it will be used as your webhook endpoint.

Documentation

For more details on how to use and extend the bot, refer to the inline documentation within the code and the official aiogram and google-generativeai libraries documentation.

Contributing

Contributions are welcome. Please submit pull requests or create issues for any features or fixes.

License

This project is open-sourced under the MIT License.