Make sure Python is installed on your operating system, then install the required dependencies using pip:

pip install --upgrade pillow mss requests

• Download and install LM Studio for your operating system
• LM Studio will be used to fetch models and run a local API server

Pick a Vision-Language Model (VLM) from LM Studio's catalog:

• Recommended: Gemma 3 12b QAT (requires 7GB of VRAM)
• Lower VRAM alternatives: Qwen3 VL 4B or 8B

Edit the MODEL parameter at the top of the script to match the exact name of the model you chose.

1. Open LM Studio
2. Navigate to the Developer tab
3. Toggle Start server (located at the top left of the screen)

Open your terminal, navigate to the directory containing the script, and run:

python TwitchChat.py

You're all set! The program should now be running.