Model comparison

Compare models or reasoning efforts by score, evaluation time and estimated cost.

Measured results

Differences are relative to the first column
Tested configuration
Effortdefault
Effort
Quick select
These models have no shared effort level

MiniMax-M3.1-Flash-PreviewFrontend −23 pts · Backend +14 pts

Metricspace-bunny · defaultMiniMax-M3.1-Flash-Preview · max
Overall
68.80
68.45−0.35 pts
Backend engineering
61
75+14 pts
Frontend interaction
80
57−23 pts
Knowledge & reasoning
75
75Tied
Visual aesthetics
59
70+11 pts
Maintainability
55
53−2 pts

Cost & time

Complete evaluation
Metricspace-bunny · defaultMiniMax-M3.1-Flash-Preview · max
Reference cost
$0.000
N/A
Evaluation time
62m 35s
54m 7s−13.5%

Dimension breakdown

Expand to inspect each subaxis.

Backend engineeringSubscores
Scoring dimensionspace-bunnydefaultMiniMax-M3.1-Flash-Previewmax
Contracts and boundaries9 / 2016 / 20
Retries and failures15 / 2019 / 20
CI and adversarial review7 / 2010 / 20
Transactions and consistency14 / 2014 / 20
Caching and regression16 / 2016 / 20
Frontend interactionSubscores
Scoring dimensionspace-bunnydefaultMiniMax-M3.1-Flash-Previewmax
Layout and adaptation18 / 2015 / 20
Interaction behavior20 / 2015 / 20
State continuity11 / 207 / 20
Accessibility and feedback12 / 157 / 15
State validation and regression19 / 2513 / 25
Knowledge & reasoningSubscores
Scoring dimensionspace-bunnydefaultMiniMax-M3.1-Flash-Previewmax
Mathematical reasoning5 / 1510 / 15
Computer science & AI15 / 1515 / 15
Science and engineering5 / 105 / 10
Life sciences and medicine20 / 2015 / 20
Humanities and general knowledge30 / 4030 / 40
Visual aestheticsSubscores
Scoring dimensionspace-bunnydefaultMiniMax-M3.1-Flash-Previewmax
Composition16 / 2020 / 20
Spacing and alignment16 / 2016 / 20
Typography and expression16 / 2017 / 20
Color and surfaces20 / 2019 / 20
Control styling16 / 2017 / 20

Subscores retain their original scale; the axis score uses a consistent conversion.

MaintainabilitySubscores
Scoring dimensionspace-bunnydefaultMiniMax-M3.1-Flash-Previewmax
Semantic clarity15 / 2016 / 20
Control-flow organization17 / 2020 / 20
Locality of changes17 / 2014 / 20
State and lifecycle16 / 2015 / 20
Implementation economy17 / 2016 / 20

Subscores retain their original scale; the axis score uses a consistent conversion.