I Let 3 AIs Compete to Build the Same App…

Today I pitted three buzzy AI coding platforms—Blitzy, Devin, and Factory AI—against the exact same real-world task (an AWS Card Demo app). With a single prompt in hand, I tracked how each tool performed on SWE Bench metrics and how much elbow grease I had to add as a developer.

From one-click magic to a few rounds of prompt tinkering, the showdown reveals which AI cranks out production-ready code, who needs more guidance, and where each platform shines or stumbles. Timestamps walk you through the demo setup, prompt engineering, benchmark results, and side-by-side code-offs so you can see exactly what happened.

Watch on YouTube