CONTRA LABS / Field Note
Pretty isn’t the same as right: One image model runs away with brief fidelity
Loading the complete research…
Related research
Six image models made real ads. Each broke differently.
We scaled the Human Creativity Benchmark into a standing benchmark, starting with ad images: six models, three client campaigns, three phases each, judged head to head by professional creatives. Meta Muse Image won the set, and every model showed a signature failure.
Contra LabsRead article
GPT Image 2 won 41.9% of logo tournaments, and still only half its logos were client-ready
10 brand designers judged GPT Image 2, Nano Banana Pro, MAI Image 2.5, and Meta Muse across 12 logo briefs. The winner is clear, and still only half its logos cleared the client bar.
Contra LabsRead article
No model owns “aesthetics”: What 8,000 designer ratings tell us about taste in image models
Five designers, five criteria, 8,000 blind pairwise ratings. Three of the four frontier models land within a few points of each other, and the best model changes depending on which dimension of visual quality you care about.
Contra LabsRead article