Bench leaderboard
This table ranks models on markup, not on whether the output works.
26 of 139 graded runs pass every execution check (19%). see verified examples →
26 of 139 graded runs pass every execution check (19%). see verified examples →
Auto-scored matrix across 198 runs · 2 models · 21 skill kinds. Scoring rubric: validity + fidelity + structure + depth + cleanliness (max 10/step).
Score over time
Dots = individual steps. Solid lines = rolling-3 average per model. As we swap models + skills the line tells the story of where we got better — or worse.
qwen-image (464) unknown (0)
Axis breakdown over time
Each rubric axis (max 2) plotted independently. Useful for diagnosing — e.g. fidelity climbs when subject-token recall improves, structure climbs when models start emitting proper semantic tags.
validity fidelity structure depth cleanliness
Model × skill (avg of 198 runs)
| Model | code-app | content-piece | dashboard | email-draft | email-sequence | image-gen | logo-svg | manifest-skill | marketing-plan | ops-doc | project-dna | research | seo-meta | slide-deck | social-card | style-guide | task-board | video-gen | web-research | website | widget | overall |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| qwen-image | 6.0 /10 (27) | 0.0 /10 (3) | 0.0 /10 (6) | 10.0 /10 (6) | 0.0 /10 (2) | 7.5 /10 (215) | 0.0 /10 (2) | 0.0 /10 (31) | 6.5 /10 (25) | 0.0 /10 (4) | 0.0 /10 (11) | 10.0 /10 (16) | 0.0 /10 (2) | 10.0 /10 (2) | 0.0 /10 (4) | 0.0 /10 (2) | 10.0 /10 (2) | 2.0 /10 (6) | 0.0 /10 (15) | 8.4 /10 (80) | 0.0 /10 (3) | 6.2 /10 |
| unknown | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — |
All graded runs
r-1786985666467-wm46r 25 / 30 qwen-image r-1786954949116-jks0t 11 / 20 qwen-image r-1786954729083-c6m4s 20 / 40 qwen-image r-1786952331083-a7811 8 / 10 qwen-image r-1786952213541-4g68i 0 / 10 qwen-image r-1786952175346-mt3je 0 / 10 qwen-image r-1786952147606-txees 0 / 10 qwen-image r-1786951615897-2hdrn 40 / 60 qwen-image r-1786951534321-hrvro 10 / 20 qwen-image r-1786941201716-yk2qr 20 / 10 qwen-image r-1786941152635-t9xsf 20 / 10 qwen-image r-1786941106684-rmmse 20 / 10 qwen-image r-1786941046131-pphe2 20 / 10 qwen-image r-1786940967479-o3xkq 20 / 10 qwen-image r-1786940908795-leh42 20 / 10 qwen-image r-1786940876359-ob1hw 2 / 10 qwen-image r-1786940813731-2635n 20 / 10 qwen-image r-1786940735522-xu5sj 10 / 10 qwen-image r-1786940600718-kue62 10 / 20 qwen-image r-1786940460821-pt40m 10 / 20 qwen-image r-1786921293615-qhtrd 19 / 10 qwen-image r-1786921244237-xszjk 19 / 10 qwen-image r-1786921192726-fesrd 19 / 10 qwen-image r-1786921127633-xk50f 19 / 10 qwen-image r-1786921069276-enhx6 19 / 10 qwen-image r-1786920809522-ezho5 19 / 30 qwen-image r-1786920770384-p4y46 2 / 10 qwen-image r-1786920721048-64acl 20 / 10 qwen-image r-1786920653240-sqb1m 10 / 10 qwen-image r-1786920513056-jk057 10 / 20 qwen-image r-1786920323057-3kbc8 10 / 20 qwen-image r-1786920237561-r826h 19 / 10 qwen-image r-1786920188296-ymwce 19 / 10 qwen-image r-1786920145620-se01y 19 / 10 qwen-image r-1786919478210-v7k96 19 / 10 qwen-image r-1786919420143-jyjig 19 / 10 qwen-image r-1786919164713-zgjr8 19 / 30 qwen-image r-1786919123438-ray44 2 / 10 qwen-image r-1786919076529-4vd3e 20 / 10 qwen-image r-1786919012029-xp66k 10 / 10 qwen-image r-1786918882948-24mod 10 / 20 qwen-image r-1786918748145-x08w7 10 / 20 qwen-image r-1786918668805-1w9ro 19 / 10 qwen-image r-1786918616467-gssqw 19 / 10 qwen-image r-1786918567777-4x3c4 19 / 10 qwen-image r-1786918503337-yuv07 19 / 10 qwen-image r-1786918456773-dwj8u 19 / 10 qwen-image r-1786918176587-hzodi 19 / 30 qwen-image r-1786918133539-my9o1 2 / 10 qwen-image r-1786918088176-kidoc 20 / 10 qwen-image