Lab
Model comparison
tmpzmbefdb4.jpg
Prompt held constant at generic-v1, so the model is the only variable.
4 model(s) not yet run: haiku, sonnet, opus, fable. Estimated cost $0.0468 at roughly $0.0117 each. Models already done are served from cache and cost nothing.
| field | Haiku 4.5 haiku |
Sonnet 5 sonnet |
Opus 5 opus |
Fable 5 fable |
|---|---|---|---|---|
| subject | not run | not run | not run | not run |
| subject_type | not run | not run | not run | not run |
| objects | not run | not run | not run | not run |
| scene | not run | not run | not run | not run |
| distinctive_features | not run | not run | not run | not run |
| dominant_colors | not run | not run | not run | not run |
| usable_as_evidence | not run | not run | not run | not run |
| unclear | not run | not run | not run | not run |
| recommend_reshoot | not run | not run | not run | not run |
| reshoot_reason | not run | not run | not run | not run |
| tags | not run | not run | not run | not run |
| description | not run | not run | not run | not run |
| tokens / latency / cost | — | — | — | — |