Community record · 2026-09-28
Claude Sonnet 5.5 · xhigh SVG
Simon Willison tested Claude Sonnet 5.5: the xhigh run took 41 seconds and cost about 5.74 cents, producing an SVG with a mostly correct frame, legs, pedals and chain, but a misshapen blue bicycle helmet. The max run spent 128,000 tokens thinking and failed to return an SVG.

Original catalog label: xhigh. A historical organizing label, not a score from a controlled benchmark.
Keep exploring
Explore the collection ↗16 models (Claude 3.5 / GPT-4o / Gemini 1.5 / Llama 3.1 ...)
The original experiment. The author liked pelicans, and very few ready-made pelican SVGs were available online at the time.
Timeline recordView record ↗
Gemini 1.5 Flash 001/002
On the original experiment's first day: 001 gave the pelican a gun; 002 gave it a barbell.
Timeline recordView record ↗
