Decision support
Qwen-Image-2.1 vs Cosmos3-Super-Text2Image
Compare canonical objects directly. The useful dimensions change with the object type, so model and hardware comparisons stay separate.
NVIDIA
Benchmark results
Head to head on 1 benchmark both models report
- Qwen-Image-2.11lead
- Cosmos3-Super-Text2Image0leads
Arena Text to Image
Arena rating · Higher is better
- Qwen-Image-2.1Best1227+77 ahead · Arena #21CI 1220 to 1233
- Cosmos3-Super-Text2Image1150Arena #36CI 1144 to 1156
Runs and sources
Qwen-Image-2.1- default122795% CI 1220 to 1233 · 17,883 votes
Cosmos3-Super-Text2Image- default115095% CI 1144 to 1156 · 17,450 votes
Arena Image Edit
Arena rating · Higher is better
- Qwen-Image-2.11363Arena #22CI 1357 to 1369
- Cosmos3-Super-Text2ImageNot reported
Runs and sources
Qwen-Image-2.1- default136395% CI 1357 to 1369 · 38,085 votes
Cosmos3-Super-Text2ImageNot reported
Each model's best published result is shown; "Not reported" means no matched evaluation, not a score of zero. Arena and DeepSWE ranks are the leaderboards' own; other ranks are among catalogue models with a score. Results from different harnesses or settings are not directly comparable.
Arena ratings: Arena (lmarena-ai/leaderboard-dataset), licensed CC BY 4.0. Ratings come from pairwise human votes and are only comparable within one leaderboard; the interval is Arena's 95% confidence interval. Model names are matched to the Vertical Tabs catalog.
Model specifications
API pricing · USD per 1M tokens
Capabilities and modalities
Architecture and availability
A dash means the field is not reported. Provider-specific performance and pricing may differ.