Model comparison
Compare models or reasoning efforts by score, evaluation time and estimated cost.
Available modelsSelected 2 / 3
Measured results
Differences are relative to the first column
Qwen 3.8 FlashFrontend +24 pts · Reasoning −15 pts
| Metric | deepseek-v4-pro · max | Qwen 3.8 Flash · default |
|---|---|---|
| Overall | 65.15 | 65.15Tied |
| Backend engineering | 67 | 68+1 pts |
| Frontend interaction | 63 | 87+24 pts |
| Knowledge & reasoning | 65 | 50−15 pts |
| Visual aesthetics | 70 | 73+3 pts |
| Maintainability | 59 | 57−2 pts |
Cost & time
Complete evaluation| Metric | deepseek-v4-pro · max | Qwen 3.8 Flash · default |
|---|---|---|
| Reference cost | $1.15 | $0.42−$0.74 |
| Evaluation time | 93m 15s | 134m 31s+44.3% |
Dimension breakdown
Expand to inspect each subaxis.
Backend engineeringSubscores
| Scoring dimension | deepseek-v4-promax | Qwen 3.8 Flashdefault |
|---|---|---|
| Contracts and boundaries | 9 / 20 | 8 / 20 |
| Retries and failures | 17 / 20 | 15 / 20 |
| CI and adversarial review | 12 / 20 | 13 / 20 |
| Transactions and consistency | 13 / 20 | 15 / 20 |
| Caching and regression | 16 / 20 | 17 / 20 |
Frontend interactionSubscores
| Scoring dimension | deepseek-v4-promax | Qwen 3.8 Flashdefault |
|---|---|---|
| Layout and adaptation | 17 / 20 | 20 / 20 |
| Interaction behavior | 16 / 20 | 19 / 20 |
| State continuity | 8 / 20 | 18 / 20 |
| Accessibility and feedback | 7 / 15 | 15 / 15 |
| State validation and regression | 15 / 25 | 15 / 25 |
Knowledge & reasoningSubscores
| Scoring dimension | deepseek-v4-promax | Qwen 3.8 Flashdefault |
|---|---|---|
| Mathematical reasoning | 15 / 15 | 15 / 15 |
| Computer science & AI | 10 / 15 | 10 / 15 |
| Science and engineering | 5 / 10 | 5 / 10 |
| Life sciences and medicine | 15 / 20 | 10 / 20 |
| Humanities and general knowledge | 20 / 40 | 10 / 40 |
Visual aestheticsSubscores
| Scoring dimension | deepseek-v4-promax | Qwen 3.8 Flashdefault |
|---|---|---|
| Composition | 19 / 20 | 18 / 20 |
| Spacing and alignment | 16 / 20 | 18 / 20 |
| Typography and expression | 18 / 20 | 17 / 20 |
| Color and surfaces | 19 / 20 | 20 / 20 |
| Control styling | 17 / 20 | 17 / 20 |
Subscores retain their original scale; the axis score uses a consistent conversion.
MaintainabilitySubscores
| Scoring dimension | deepseek-v4-promax | Qwen 3.8 Flashdefault |
|---|---|---|
| Semantic clarity | 15 / 20 | 15 / 20 |
| Control-flow organization | 20 / 20 | 20 / 20 |
| Locality of changes | 15 / 20 | 15 / 20 |
| State and lifecycle | 17 / 20 | 16 / 20 |
| Implementation economy | 17 / 20 | 17 / 20 |
Subscores retain their original scale; the axis score uses a consistent conversion.