Original prompt
Generate an SVG of a pelican riding a bicycle.
骑自行车的鹈鹕
4 runs · 02.A.01 · SVG
PROMPT/Phase1_TEST_PROMPTS_BILINGUAL.md · The original prompt is kept verbatim
Generate an SVG of a pelican riding a bicycle.
生成一幅鹈鹕骑自行车的 SVG 图像。
The deliverable should be a valid, standalone SVG that clearly depicts a pelican riding a bicycle. The pelican should be recognizable through characteristic features such as its long bill, throat pouch, wings, and avian body shape, while the bicycle should include essential structures such as wheels, frame, handlebars, saddle, and pedals. Their pose and contact points should communicate riding rather than merely standing near or flying over the bicycle. The composition should be complete, balanced, and crisp at different scales.
All three results use the same width and the same preview ratio; no model is marked as best.
Cross-model comparison must note differences in harness, budget and isolation.
| Metric | DeepSeek-V4.1-Flashrun-deepseek-v4-1-flash-exp-0910-task-02-r1 | Muse Spark 1.3run-muse-spark-1-3-task-02-r1 | GPT-5.6 Solrun-gpt-5-6-sol-task-02-r1 | K3run-k3-task-02-r1 |
|---|---|---|---|---|
| Runtime | 14:38 | 0:50 | 2:06 | 6:55 |
| API calls | 48 | 5 | 3 | 11 |
| Tool calls | 47 | 4 | 2 | 12 |
| Failures / retries | 3 | 0 | 0 | 0 |
| Token throughput | 1.4M | 65.2K | 87.6K | 453.1K |
| AI review | 94/100 | 93/100 | 88/100 | 94/100 |
Two runs of the same task version can be compared side by side.
Type to search; ↑ ↓ to choose, Enter to open
No matching archive records.