Model profile
o1 Mini
Evidence summary
Safety. o1 Mini has an estimated Safety rank of #94; its 90% source-sensitivity interval is #31–#237. Its behavior-only rank is #104; company governance moves the combined estimate to #94. Published Safety evidence spans 11 eval lineages and 7 of 7 components. Its strongest relative result is Large-scale Moral Machine experiment on LLMs (human_choice_distance, #1 of 39); its weakest is AIRBench 2024 Safety Scenarios (safety_scenarios, #68 of 80).
Freedom. o1 Mini has an estimated Freedom rank of #337; its 90% source-sensitivity interval is #143–#594. Published Freedom evidence spans 8 eval lineages and 1 of 1 components. Its strongest relative result is AIRBench 2024 Safety Scenarios (safety_scenarios, #13 of 80); its weakest is SpeechMap model completion (complete_pct, #161 of 181).
Compare this model
Only models sharing at least one published sub-eval are listed.
Official and reference links
- Artificial Analysis ↗o1-mini
- OpenRouter ↗openai/o1-mini
- System card ↗Family-level model document · OpenAI · first party
- Release source ↗direct research preview identity date
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 |
|---|---|---|---|---|
| AIRBench 2024 Safety Scenariossafety_scenarios | #68 / 80 | ↑0.452 | Source ↗official | |
| Confabulationsconfabulation_rate | #31 / 52 | ↓26.24 | Source ↗official | |
| Enkrypt AI Safety Leaderboardbias_attack_non_success_rate | #93 / 248 | ↑19.12 | Source ↗official | |
| Enkrypt AI Safety Leaderboardcbrn_attack_non_success_rate | #40 / 248 | ↑92.67 | Source ↗official | |
| Enkrypt AI Safety Leaderboardharmful_attack_non_success_rate | #42 / 248 | ↑92.78 | Source ↗official | |
| Enkrypt AI Safety Leaderboardtoxicity_attack_non_success_rate | #96 / 246 | ↑97.27 | Source ↗official | |
| HELM Safetyanthropic_red_team | #50 / 80 | ↑0.983 | Source ↗official | |
| HELM Safetybbq | #11 / 80 | ↑0.969 | Source ↗official | |
| HELM Safetyharmbench | #23 / 80 | ↑0.885 | Source ↗official | |
| HELM Safetysimple_safety_tests | #58 / 80 | ↑0.97 | Source ↗official | |
| HELM Safetyxstest | #24 / 80 | ↑0.97 | Source ↗official | |
| HUMAINE Trust, Ethics and Safetytrust_ethics_safety_score | #41 / 54 | ↑25.53 | Source ↗official | |
| Large-scale Moral Machine experiment on LLMshuman_choice_distance | #1 / 39 | ↓0.7105 | Source ↗official | |
| SORRY-Benchavg | #20 / 51 | ↓0.24 | 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 |
|---|---|---|---|---|
| AIRBench 2024 Safety Scenariossafety_scenarios | #13 / 80 | ↓0.452 | Source ↗official | |
| Enkrypt AI Safety Leaderboardcbrn_attack_non_success_rate | #207 / 248 | ↓92.67 | Source ↗official | |
| Enkrypt AI Safety Leaderboardharmful_attack_non_success_rate | #205 / 248 | ↓92.78 | Source ↗official | |
| Enkrypt AI Safety Leaderboardtoxicity_attack_non_success_rate | #147 / 246 | ↓97.27 | Source ↗official | |
| HELM Safetyanthropic_red_team | #28 / 80 | ↓0.983 | Source ↗official | |
| HELM Safetyharmbench | #58 / 80 | ↓0.885 | Source ↗official | |
| HELM Safetysimple_safety_tests | #20 / 80 | ↓0.97 | Source ↗official | |
| HELM Safetyxstest | #24 / 80 | ↑0.97 | Source ↗official | |
| SORRY-Benchavg | #31 / 51 | ↑0.24 | Source ↗official | |
| SpeechMap model completioncomplete_pct | #161 / 181 | ↑33.8 | Source ↗official |
Values evaluations
Descriptive values and political-framing results are separate from safety/ethics ranks. Each strip shows the evaluation’s observed model range; its endpoint labels state what lower and higher values mean.