Telegram AI bot using Ollama

Powered by Ollama and python-telegram-bot

✨ Features

Local Processing: Your data stays on your machine. No cloud API subscriptions required.
Persistent History: Automatically saves conversation history per user.
Configurable: Easily swap models, adjust context windows, or change history depth.
Type Awareness: Displays a "typing..." status while the AI is thinking.

Install Ollama from ollama.com/download.
Pull a model using ollama pull <model_name>.
- For example: ollama pull gemma3
Pip install
```
pip install -r requirements.txt
```
Rename .env.copy to .env
```
mv .env.copy .env
```
Fill in your Telegram bot token in .env

The .env file has the right bot token. Run /newbot at: t.me/botfather
The .env file has a model that you have downloaded (check with ollama ls)

You can customize the behavior by editing the .env file:

Variable	Description	Default
`TELEGRAM_BOT_TOKEN`	Your unique Bot API token.	Required
`OLLAMA_MODEL`	The model name (must be downloaded).	`gemma3`
`MAX_HISTORY`	Number of messages to remember for context.	`10`
`CONTEXT_WINDOW`	The `num_ctx` passed to Ollama.	`2048`
`KEEP_ALIVE`	How long to keep the model in memory.	`-1` (infinite)

This project is licensed under the MIT License. See the LICENSE file for more details.