README.md

Telegram Translation Bot with integrated local AI

A Telegram bot that translates messages using a locally running LLM (Ollama), so no data leaving your machine.

Example:

• Input: Hello, how are you?
• Output: Привет, как дела?

How translation works

• Language is detected using langdetect
• Target language is selected via simple rules (you can easily customise it however you like):
  • EN / DE -> Russian
  • RU -> English
• A prompt is sent to the local LLM
• The model returns only the translated text

Setup

1. Clone repository

2. Create virtual environment

python -m venv .venv
source .venv/bin/activate   # Windows: .venv\Scripts\activate

3. Install dependencies

pip install -r requirements.txt

4. Setup environment variables

Create a .env file:

TOKEN=your_telegram_bot_token
MODEL=model_of_your_choice

5. Start Ollama

Make sure Ollama is running locally:

ollama run qwen3

Default API endpoint:

http://localhost:11434/api/generate

6. Run the bot

python bot.py