Tried GPT-4o, Claude 3.5 Sonnet, and Claude 3.5 Haiku on the same image, and your tool was significantly more accurate. Could you share what model you're using? Is it a pre-trained model that you fine-tuned, or did you train your own model from scratch?
