AI Benchmark Challenges Models to Generate an SVG of a Frog With a Habsburg Jaw, Giving Each Model Three Attempts Per Month

A new personal AI benchmark challenges models to generate an SVG of a frog with a Habsburg jaw, giving each model three attempts per month. The benchmark's first run, completed on Aug 2, 2026, took 64.0 seconds and produced a 3,900-byte SVG. The annotations were mostly structural labels but included editorializing about the jaw feature, describing it as a "massive protruding mandible, with the upper lip "recessed, tucked behind the jaw" and lower teeth "protruding" over the upper lip—offering anatomical interpretation beyond plain labeling.

The second run, also on Aug 2, 2026, took 42.0 seconds and produced a 3,465-byte SVG. Its annotations used structural labels but included editorializing on exaggerated anatomy, such as a "HUGE protruding Habsburg jaw" and "lower teeth jutting over the upper lip, plus implied royal bearing and mood via "droopy regal eyelids. The third run, on the same date, took 59.6 seconds and produced a 3,909-byte SVG."