oneshotlm
Compare all models on this prompt →

Pelican on a bicycle · qwen3.6-35b-a3b

Generate an inline SVG illustration of a pelican riding a bicycle.

⛨ sandboxed Open in new tab ↗
transcript · 3 messages
assistant
thinking
The user wants me to create a single-page web app with an inline SVG illustration of a pelican riding a bicycle, all in one self-contained index.html file. Let me create something visually appealing with SVG.

Let me design a pelican on a bicycle - I'll need to think about the SVG shapes:
- Bicycle: two wheels, frame, handlebars, seat
- Pelican: body, long beak/pouch, head, legs pedaling

Let me create a nice illustration.
write /workspace/index.html (16922 bytes)
toolResult

Successfully wrote 16922 bytes to /workspace/index.html

assistant
thinking
The file has been written. Let me verify it exists and provide a summary.

Done! I've created `/workspace/index.html` — a single self-contained file featuring an inline SVG illustration of a pelican riding a bicycle. Here's what's in it: **The SVG scene includes:** - 🐦 A white pelican with an orange beak & pouch, wearing a **sunhat with a pink band** and a **pink scarf** - 🚲 A red-framed bicycle with detailed spoked wheels, fenders, chain, pedals, handlebars & seat - 🛣️ A road with dashed lane markings, grassy ground with tiny flowers, and a sky gradient with sun **Animations (CSS-based, no external resources):** - Spinning wheels & pedals - Gentle bird body bobbing - Wing flapping - Eye blinking - Beak swaying - Scrolling clouds with parallax on mouse move **Fully offline** — zero network requests, no CDNs, pure inline CSS + JS + SVG.

TAKE IT TO CODE

A single self-contained HTML file — copy it and open in any browser, or reuse the prompt with your own pi-agent run.

RUN

modelqwen3.6-35b-a3b
vendorQwen
harnesspi-agent
generated2026-07-31

EVALUATION

3/5 working

A simple animated SVG pelican pedaling a bicycle across a scenic background, with pedaling motion but no interactivity.

Vision grade · anthropic/claude-sonnet-5

MEASUREMENTS

Raw readings from the capture, not a verdict — the grade above is the only judgement. Percentages are how much of the screen changed; the measure under-reports sparse motion on a dark background.

idle motion2.3%
on click2.4%
on drag2.3%
on wheel1.4%
on enter + space1.8%
on arrow keys3.4%
on w/a/s/d3.2%
frame spread40.1 / 255
console errors0
js errors none

METRICS

runtime234.2s
tokens in11k
tokens out8.1k
cached2.1k
cost / run$0.0094