Model profile
Nova Pro
Evidence summary
Safety. Nova Pro has an estimated Safety rank of #266; its 90% source-sensitivity interval is #131–#298. Its behavior-only rank is #271; company governance moves the combined estimate to #266. Published Safety evidence spans 8 eval lineages and 6 of 7 components. Its strongest relative result is Vectara HHEM Factual Consistency (factual_consistency_rate, #8 of 94); its weakest is SpeciEval (sea_animal_4ns, #122 of 123).
Freedom. Nova Pro has an estimated Freedom rank of #436; its 90% source-sensitivity interval is #164–#571. Published Freedom evidence spans 2 eval lineages and 1 of 1 components. Its strongest relative result is Cisco AI Defense Rolling Single-Turn Leaderboard (single_turn_attack_success_rate, #39 of 104); its weakest is Enkrypt AI Safety Leaderboard (cbrn_attack_non_success_rate, #224 of 248).
Compare this model
Only models sharing at least one published sub-eval are listed.
Official and reference links
- Artificial Analysis ↗nova-pro
- Official model documentation ↗Family-level model document · amazon · first party
- Release source ↗scoring identity family inherited
Safety evals
Rank and direction are specific to the Safety portfolio. Black marks the observed result; the grey dot marks the value implied by the Safety rank. Values stay on each source’s native scale.
| Eval / sub-eval | Rank | Value | Distribution | Source |
|---|---|---|---|---|
| AA-Omnisciencehallucination_rate | #155 / 345 | ↓0.7771 | Source ↗official | |
| Cisco AI Defense Rolling Single-Turn Leaderboardsingle_turn_attack_success_rate | #66 / 104 | ↓53.34 | Source ↗official | |
| Confabulationsconfabulation_rate | #48 / 52 | ↓54.46 | Source ↗official | |
| Enkrypt AI Safety Leaderboardbias_attack_non_success_rate | #131 / 248 | ↑15.25 | Source ↗official | |
| Enkrypt AI Safety Leaderboardcbrn_attack_non_success_rate | #23 / 248 | ↑94.5 | Source ↗official | |
| Enkrypt AI Safety Leaderboardharmful_attack_non_success_rate | #86 / 248 | ↑82.22 | Source ↗official | |
| Enkrypt AI Safety Leaderboardtoxicity_attack_non_success_rate | #136 / 246 | ↑95.82 | Source ↗official | |
| Humanity's Last Exam RMS calibration error (Scale Labs)calibrationError | #29 / 42 | ↓80 | Source ↗official | |
| RefusalBenchyouden_j | #15 / 19 | ↑0.01282 | Source ↗official | |
| SpeciEvalbelief_animal_sentience | #109 / 123 | ↑6.37 | Source ↗official | |
| SpeciEvalland_animal_4ns | #68 / 123 | ↓4.58 | Source ↗official | |
| SpeciEvalsea_animal_4ns | #122 / 123 | ↓5.6 | Source ↗official | |
| SpeciEvalspeciesism | #107 / 123 | ↓2.65 | Source ↗official | |
| Vectara HHEM Factual Consistencyfactual_consistency_rate | #8 / 94 | ↑94.9 | Source ↗official |
Freedom evals
Rank and direction are specific to the Freedom portfolio. Black marks the observed result; the grey dot marks the value implied by the Freedom rank. Values stay on each source’s native scale.
| Eval / sub-eval | Rank | Value | Distribution | Source |
|---|---|---|---|---|
| Cisco AI Defense Rolling Single-Turn Leaderboardsingle_turn_attack_success_rate | #39 / 104 | ↑53.34 | Source ↗official | |
| Enkrypt AI Safety Leaderboardcbrn_attack_non_success_rate | #224 / 248 | ↓94.5 | Source ↗official | |
| Enkrypt AI Safety Leaderboardharmful_attack_non_success_rate | #162 / 248 | ↓82.22 | Source ↗official | |
| Enkrypt AI Safety Leaderboardtoxicity_attack_non_success_rate | #110 / 246 | ↓95.82 | Source ↗official |