Model profile
GPT 4.1 Mini
OpenAIdeveloper
2025-04-14release date
#121 / 309overall rank
19eval lineages
Evidence summary
GPT 4.1 Mini has an estimated overall rank of #121; its 90% source-sensitivity interval is #25–#214. Its behavior-only rank is #133; company governance moves the combined estimate to #121. Published evidence spans 19 evals and 7 of 7 behavior components. Its strongest relative result is HELM Safety (simple_safety_tests, #1 of 80); its weakest is LiveSecBench (factuality, #40 of 43).
Compare this model
Only models sharing at least one published sub-eval are listed.
Official and reference links
- Artificial Analysis ↗gpt-4-1-mini
- OpenRouter ↗openai/gpt-4.1-mini
- Official model documentation ↗Family-level model document · openai · first party
- Release source ↗direct research preview identity date
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 | Source |
|---|---|---|---|---|
| AA-Omnisciencehallucination_rate | #285 / 328 | ↓0.9272 | Source ↗official | |
| AgentDrive Safety Compliancescr | #2 / 48 | ↑96.25 | Source ↗official | |
| AIRBench 2024 Safety Scenariossafety_scenarios | #50 / 80 | ↑0.604 | Source ↗official | |
| BlueBench AttaQ-100attaq_harmlessness_reward_pct | #8 / 18 | ↑85.96 | Source ↗official | |
| Cisco AI Defense Rolling Single-Turn Leaderboardsingle_turn_attack_success_rate | #92 / 104 | ↓72.73 | Source ↗official | |
| Contextual MoralChoicehuman_agreement | #12 / 22 | ↑0.4 | Source ↗official | |
| Enkrypt AI Safety Leaderboardbias_attack_non_success_rate | #123 / 241 | ↑15.5 | Source ↗official | |
| Enkrypt AI Safety Leaderboardcbrn_attack_non_success_rate | #110 / 241 | ↑88.67 | Source ↗official | |
| Enkrypt AI Safety Leaderboardharmful_attack_non_success_rate | #147 / 241 | ↑58.33 | Source ↗official | |
| Enkrypt AI Safety Leaderboardtoxicity_attack_non_success_rate | #148 / 239 | ↑95.23 | Source ↗official | |
| FORTRESSaverage_risk_score | #43 / 49 | ↓59.19 | Source ↗official | |
| FORTRESSover_refusal_score | #6 / 46 | ↓1.38 | Source ↗official | |
| HELM Safetyanthropic_red_team | #27 / 80 | ↑0.993 | Source ↗official | |
| HELM Safetybbq | #47 / 80 | ↑0.921 | Source ↗official | |
| HELM Safetyharmbench | #27 / 80 | ↑0.856 | Source ↗official | |
| HELM Safetysimple_safety_tests | #1 / 80 | ↑1 | Source ↗official | |
| HELM Safetyxstest | #17 / 80 | ↑0.974 | Source ↗official | |
| LiveSecBenchethics | #35 / 43 | ↑23.73 | Source ↗official | |
| LiveSecBenchfactuality | #40 / 43 | ↑22.2 | Source ↗official | |
| LiveSecBenchlegality | #35 / 43 | ↑19.69 | Source ↗official | |
| LiveSecBenchprivacy | #39 / 43 | ↑18.39 | Source ↗official | |
| LiveSecBenchpsychological_health | #32 / 43 | ↑30.95 | Source ↗official | |
| MASKlying_probability_pct | #40 / 53 | ↓50 | Source ↗official | |
| MonitoringBench Full-Trajectory Monitorfull_trajectory_catch_rate_at_1pct_fpr_percent | #12 / 13 | ↑4 | Source ↗official | |
| PHAREbias_resistance_diagnostic | #1 / 66 | ↑0.8812 | Source ↗official | |
| PHAREhallucination_resistance_diagnostic | #52 / 70 | ↑0.6781 | Source ↗official | |
| PHAREharm_resistance_diagnostic | #61 / 70 | ↑0.8339 | Source ↗official | |
| PHAREjailbreak_resistance_diagnostic | #47 / 67 | ↑0.4255 | Source ↗official | |
| TACbase_welfare_rate | #28 / 76 | ↑29.49 | Source ↗self run | |
| UAVBench safety-critical decision recognitionethical_safety_critical_accuracy | #17 / 27 | ↑0.67 | Source ↗official |