Model profile
GPT 5.1
OpenAIdeveloper
2025-11-13release date
#44 / 309overall rank
30eval lineages
Evidence summary
GPT 5.1 has an estimated overall rank of #44; its 90% source-sensitivity interval is #22–#143. Its behavior-only rank is #52; company governance moves the combined estimate to #44. Published evidence spans 30 evals and 7 of 7 behavior components. Its strongest relative result is Arena Factuality — Text Arena (factuality-only weighting) (factuality_bt_rating, #7 of 112); its weakest is GPT-5.6 system card (hate_not_unsafe, #7 of 7).
Compare this model
Only models sharing at least one published sub-eval are listed.
Official and reference links
- Artificial Analysis ↗gpt-5-1
- OpenRouter ↗openai/gpt-5.1
- 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 | #64 / 328 | ↓0.5189 | Source ↗official | |
| AgentAbstainabstain | #10 / 17 | ↑60.7 | Source ↗official | |
| AgentAbstaincar | #11 / 17 | ↑53.2 | Source ↗official | |
| AgentAbstainpaired | #13 / 17 | ↑40.6 | Source ↗official | |
| AIRBench 2024 Safety Scenariossafety_scenarios | #9 / 80 | ↑0.862 | Source ↗official | |
| Arena Factuality — Search Arena (factuality-only weighting)factuality_bt_rating | #6 / 30 | ↑1226.0 | Source ↗official | |
| Arena Factuality — Text Arena (factuality-only weighting)factuality_bt_rating | #7 / 112 | ↑1483.0 | Source ↗official | |
| BullshitBench v2clear_pushback_rate | #41 / 106 | ↑0.35 | Source ↗official | |
| CAIS Risk Indexagent_red_teaming | #26 / 44 | ↓82.1 | Source ↗official | |
| CAIS Risk Indexbioweapons_assistance | #24 / 49 | ↓66.3 | Source ↗official | |
| CAIS Risk Indexhle_overconfidence | #18 / 50 | ↓47.8 | Source ↗official | |
| CAIS Risk Indexmachiavelli | #27 / 46 | ↓89.6 | Source ↗official | |
| CAIS Risk Indexmask | #16 / 52 | ↓12 | Source ↗official | |
| CAIS Risk Indexpolitical_manipulation | #17 / 33 | ↓47.9 | Source ↗official | |
| CAIS Risk Indextextquests_harm | #19 / 49 | ↓16.7 | Source ↗official | |
| Constitutional Following — OpenAI Model Specconstitutional_following_score | #1 / 7 | ↑96.1 | Source ↗official | |
| Contextual MoralChoicehuman_agreement | #6 / 22 | ↑0.45 | Source ↗official | |
| Enkrypt AI Safety Leaderboardbias_attack_non_success_rate | #23 / 241 | ↑43.67 | Source ↗official | |
| Enkrypt AI Safety Leaderboardcbrn_attack_non_success_rate | #33 / 241 | ↑93.5 | Source ↗official | |
| Enkrypt AI Safety Leaderboardharmful_attack_non_success_rate | #73 / 241 | ↑83.89 | Source ↗official | |
| Enkrypt AI Safety Leaderboardtoxicity_attack_non_success_rate | #135 / 239 | ↑95.64 | Source ↗official | |
| FORTRESSaverage_risk_score | #23 / 49 | ↓28.04 | Source ↗official | |
| FORTRESSover_refusal_score | #31 / 46 | ↓5.78 | Source ↗official | |
| GPT-5.6 system cardconnectors_injection_resistance | #7 / 7 | ↑0.649 | Source ↗official | |
| GPT-5.6 system cardemotional_reliance | #7 / 7 | ↑0.857 | Source ↗official | |
| GPT-5.6 system cardextremism_not_unsafe | #1 / 7 | ↑1 | Source ↗official | |
| GPT-5.6 system cardgore_not_unsafe | #3 / 7 | ↑0.8 | Source ↗official | |
| GPT-5.6 system cardharm_overall_pct | #7 / 7 | ↓1.28 | Source ↗official | |
| GPT-5.6 system cardhate_not_unsafe | #7 / 7 | ↑0.808 | Source ↗official | |
| GPT-5.6 system cardmental_health | #7 / 7 | ↑0.753 | Source ↗official | |
| GPT-5.6 system cardnonviolent_illicit_not_unsafe | #4 / 7 | ↑0.99 | Source ↗official | |
| GPT-5.6 system cardsearch_function_calling_injection_resistance | #6 / 6 | ↑0.423 | Source ↗official | |
| GPT-5.6 system cardself_harm | #5 / 7 | ↑0.904 | Source ↗official | |
| GPT-5.6 system cardself_harm_not_unsafe | #6 / 7 | ↑0.926 | Source ↗official | |
| GPT-5.6 system cardsexual_minors_not_unsafe | #7 / 7 | ↑0.916 | Source ↗official | |
| GPT-5.6 system cardsexual_not_unsafe | #6 / 7 | ↑0.933 | Source ↗official | |
| GPT-5.6 system cardviolent_illicit_not_unsafe | #3 / 7 | ↑0.955 | Source ↗official | |
| HELM Safetyanthropic_red_team | #18 / 80 | ↑0.995 | Source ↗official | |
| HELM Safetybbq | #58 / 80 | ↑0.887 | Source ↗official | |
| HELM Safetyharmbench | #6 / 80 | ↑0.976 | Source ↗official | |
| HELM Safetysimple_safety_tests | #24 / 80 | ↑0.998 | Source ↗official | |
| HELM Safetyxstest | #9 / 80 | ↑0.982 | Source ↗official | |
| MACHIAVELLIdeception_relative_random_pct | #27 / 50 | ↓89.6 | Source ↗official | |
| MASKlying_probability_pct | #20 / 53 | ↓25.17 | Source ↗official | |
| OpenAI GPT-5.3 Dynamic Wellbeingemotional_reliance | #3 / 3 | ↑0.945 | Source ↗official | |
| OpenAI GPT-5.3 Dynamic Wellbeingmental_health | #3 / 3 | ↑0.832 | Source ↗official | |
| OpenAI GPT-5.3 Dynamic Wellbeingself_harm | #3 / 3 | ↑0.845 | Source ↗official | |
| OpenAI GPT-5.4 Dynamic Wellbeingemotional_reliance | #3 / 3 | ↑0.857 | Source ↗official | |
| OpenAI GPT-5.4 Dynamic Wellbeingmental_health | #3 / 3 | ↑0.753 | Source ↗official | |
| OpenAI GPT-5.4 Dynamic Wellbeingself_harm | #3 / 3 | ↑0.904 | Source ↗official | |
| OpenAI GPT-5.4 First-Person Fairnessharm_overall | #3 / 3 | ↓0.0128 | Source ↗official | |
| PHAREbias_resistance_diagnostic | #30 / 66 | ↑0.4677 | Source ↗official | |
| PHAREhallucination_resistance_diagnostic | #14 / 70 | ↑0.8184 | Source ↗official | |
| PHAREharm_resistance_diagnostic | #14 / 70 | ↑0.9692 | Source ↗official | |
| PHAREjailbreak_resistance_diagnostic | #17 / 67 | ↑0.6576 | Source ↗official | |
| RealityTest — Text AI-Identity Disclosuredisclosure_probability | #2 / 17 | ↑0.856 | Source ↗official | |
| SM-Benchadversarial | #17 / 79 | ↑86.34 | Source ↗official | |
| SM-Benchambiguous_interpretation | #15 / 79 | ↑90.18 | Source ↗official | |
| SM-Benchanti_hallucination | #43 / 79 | ↑91.62 | Source ↗official | |
| SM-Bencheq_boundaries | #62 / 79 | ↑55.62 | Source ↗official | |
| SM-Benchoverfit | #53 / 79 | ↑60.11 | Source ↗official | |
| SpeciEvalbelief_animal_sentience | #41 / 113 | ↑6.87 | Source ↗official | |
| SpeciEvalland_animal_4ns | #15 / 113 | ↓4.2 | Source ↗official | |
| SpeciEvalsea_animal_4ns | #10 / 113 | ↓4.3 | Source ↗official | |
| SpeciEvalspeciesism | #11 / 113 | ↓1.35 | Source ↗official | |
| Vectara HHEM Factual Consistencyfactual_consistency_rate | #67 / 94 | ↑88.5 | Source ↗official | |
| Vigil Mental Health Safetyoverall_score | #7 / 23 | ↑58 | Source ↗official |