🔬 Multi-Model Comparison

The original study tested only GPT-4 (text-only, pre-vision). As multimodal AI evolves rapidly, this tool tracks how different models — including vision-capable ones — perform against GIF-based CAPTCHAs. Add your own test results or explore the pre-loaded data.

📋 Model Selector

🏆 Model Scorecards

📊 Comparison Matrix

📈 Capability Radar

📅 AI Vision Timeline

➕ Add Model Results