Model profile
O3 Pro
2eval lineages
Evidence summary
Published evidence spans 2 evals and 1 of 7 behavior components. Its strongest relative result is MASK (lying_probability_pct, #15 of 53); its weakest is Confabulations (confabulation_rate, #25 of 52).
Compare this model
Only models sharing at least one published sub-eval are listed.
Official and reference links
- Artificial Analysis ↗o3-pro
- OpenRouter ↗openai/o3-pro
- Official model documentation ↗Family-level model document · openai · first party
Published eval results
Rank is within that sub-eval. Black marks the observed result; the grey dot marks the value implied by the global rank. Values stay on each source’s native scale.
| Eval / sub-eval | Rank | Value | Distribution | Better | Source |
|---|---|---|---|---|---|
| Confabulationsconfabulation_rate | #25 / 52 | 23.38 | ↓ lower | Source ↗official | |
| MASKlying_probability_pct | #15 / 53 | 17.5 | ↓ lower | Source ↗official |