Imagen 4.0 on DTP Benchmark
At a glance
Good enough on 0/35 tasks at the 95% bar. Cheapest qualifier on 0 tasks.
Cost vs quality across all tasks
qualifies at 90% bar · doesn't qualify · ★ this model is the best on that task. Lower + further right = cheaper + higher quality. Y-axis is log-scaled.
Per-task breakdown
| Category | Task | Quality | Confidence | Cost / call | vs best | Qualifies @ 90% |
|---|---|---|---|---|---|---|
| Infrastructure & Utility | report_image_generation autogenerated | 3.70 | low · n=1 | $0.040000 | — | no |