Model comparison
Selected 2 / 5See the differences. Choose what matters to you.
Available modelsSelected 2 / 5
Differences vs. gpt-6-astra · high
Score, cost & time differences
Current baseline · gpt-6-astra · high
96Frontend interaction
Compared with gpt-6-astra · high
- deepseek-v4.1-flash-1017 · max
Frontend interaction: tiedReference cost $0.062 ($0.66 less)Evaluation time 21m 0s (0m 14s less)
Differences reflect these evaluation tasks. Cost and time cover the same dimension, not a typical conversation. Scoring & cost notes
| Metric | gpt-6-astra · high | deepseek-v4.1-flash-1017 · max |
|---|---|---|
| Frontend interaction | gpt-6-astrahigh 96 | deepseek-v4.1-flash-1017max 96Tied |
| Overall | gpt-6-astrahigh 84.85 | deepseek-v4.1-flash-1017max 74.30−10.55 pts |
| Backend engineering | gpt-6-astrahigh 92 | deepseek-v4.1-flash-1017max 72−20 pts |
| Knowledge & Reasoning | gpt-6-astrahigh 85 | deepseek-v4.1-flash-1017max 75−10 pts |
| Visual aesthetics | gpt-6-astrahigh 70 | deepseek-v4.1-flash-1017max 61−9 pts |
| Maintainability | gpt-6-astrahigh 70 | deepseek-v4.1-flash-1017max 53−17 pts |
Frontend interaction evaluation · Evaluation consumption, not a monthly subscription fee
Cost & time
Frontend interaction evaluation| Metric | gpt-6-astra · high | deepseek-v4.1-flash-1017 · max |
|---|---|---|
| Reference cost | gpt-6-astrahigh $0.72 | deepseek-v4.1-flash-1017max $0.062−$0.66 |
| Evaluation time | gpt-6-astrahigh 21m 14s | deepseek-v4.1-flash-1017max 21m 0s−1.1% |
Capability profiles
Show up to five models. Select names to focus the chart.
Behind each score
Expand to inspect each subaxis.
Backend engineeringSubscores
| Scoring dimension | gpt-6-astrahigh | deepseek-v4.1-flash-1017max |
|---|---|---|
| Contracts and boundaries | 19 / 20 | 17 / 20 |
| Retries and failures | 19 / 20 | 15 / 20 |
| CI and adversarial review | 16 / 20 | 11 / 20 |
| Transactions and consistency | 18 / 20 | 14 / 20 |
| Caching and regression | 20 / 20 | 15 / 20 |
Frontend interactionSubscores
| Scoring dimension | gpt-6-astrahigh | deepseek-v4.1-flash-1017max |
|---|---|---|
| Layout and adaptation | 20 / 20 | 20 / 20 |
| Interaction behavior | 20 / 20 | 20 / 20 |
| State continuity | 20 / 20 | 20 / 20 |
| Accessibility and feedback | 15 / 15 | 11 / 15 |
| State validation and regression | 21 / 25 | 25 / 25 |
Knowledge & ReasoningSubscores
| Scoring dimension | gpt-6-astrahigh | deepseek-v4.1-flash-1017max |
|---|---|---|
| Mathematical reasoning | 15 / 15 | 15 / 15 |
| Computer science & AI | 15 / 15 | 15 / 15 |
| Science and engineering | 10 / 10 | 5 / 10 |
| Life sciences and medicine | 15 / 20 | 15 / 20 |
| Humanities and general knowledge | 30 / 40 | 25 / 40 |
Visual aestheticsSubscores
| Scoring dimension | gpt-6-astrahigh | deepseek-v4.1-flash-1017max |
|---|---|---|
| Composition | 19 / 20 | 19 / 20 |
| Spacing and alignment | 15 / 20 | 15 / 20 |
| Typography and expression | 19 / 20 | 16 / 20 |
| Color and surfaces | 19 / 20 | 19 / 20 |
| Control styling | 17 / 20 | 16 / 20 |
Subscores retain their original scale; the axis score uses a consistent conversion.
MaintainabilitySubscores
| Scoring dimension | gpt-6-astrahigh | deepseek-v4.1-flash-1017max |
|---|---|---|
| Semantic clarity | 16 / 20 | 15 / 20 |
| Control-flow organization | 20 / 20 | 19 / 20 |
| Locality of changes | 17 / 20 | 14 / 20 |
| State and lifecycle | 18 / 20 | 17 / 20 |
| Implementation economy | 18 / 20 | 16 / 20 |
Subscores retain their original scale; the axis score uses a consistent conversion.