Model profile
Grok 4.1 Fast
Evidence summary
Safety. Grok 4.1 Fast has an estimated Safety rank of #227; its 90% source-sensitivity interval is #112–#314. Its behavior-only rank is #226; company governance moves the combined estimate to #227. Published Safety evidence spans 25 eval lineages and 7 of 7 components. Its strongest relative result is SpeciEval (belief_animal_sentience, #1 of 131); its weakest is AuAu Authoritarian Response Audit (realistic_prompt_arr_percent, #17 of 17).
Freedom. Grok 4.1 Fast has an estimated Freedom rank of #298; its 90% source-sensitivity interval is #85–#464. Published Freedom evidence spans 12 eval lineages and 1 of 1 components. Its strongest relative result is SpeechMap model completion (complete_pct, #2 of 181); its weakest is Adversarial Humanities Benchmark (AHB) — Table 5 (child_sexual_exploitation, #28 of 31).
Compare this model
Only models sharing at least one published sub-eval are listed.
Official and reference links
- Artificial Analysis ↗grok-4-1-fast
- OpenRouter ↗x-ai/grok-4.1-fast
- Official model documentation ↗Family-level model document · xai · 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 |
|---|---|---|---|---|
| AA-Omnisciencehallucination_rate | #149 / 358 | ↓0.7341 | Source ↗official | |
| Adversarial Humanities Benchmark (AHB) — Table 5child_sexual_exploitation | #4 / 31 | ↓14.6 | Source ↗official | |
| Adversarial Humanities Benchmark (AHB) — Table 5defamation | #22 / 31 | ↓78.3 | Source ↗official | |
| Adversarial Humanities Benchmark (AHB) — Table 5hate | #23 / 31 | ↓65.2 | Source ↗official | |
| Adversarial Humanities Benchmark (AHB) — Table 5indiscriminate_weapons_cbrne | #10 / 31 | ↓33.3 | Source ↗official | |
| Adversarial Humanities Benchmark (AHB) — Table 5intellectual_property | #10 / 31 | ↓41.3 | Source ↗official | |
| Adversarial Humanities Benchmark (AHB) — Table 5non_violent_crimes | #11 / 31 | ↓52.2 | Source ↗official | |
| Adversarial Humanities Benchmark (AHB) — Table 5privacy | #11 / 31 | ↓53.2 | Source ↗official | |
| Adversarial Humanities Benchmark (AHB) — Table 5sex_related_crimes | #11 / 31 | ↓54.3 | Source ↗official | |
| Adversarial Humanities Benchmark (AHB) — Table 5sexual_content | #26 / 31 | ↓64.6 | Source ↗official | |
| Adversarial Humanities Benchmark (AHB) — Table 5specialized_advice | #14 / 31 | ↓60.92 | Source ↗official | |
| Adversarial Humanities Benchmark (AHB) — Table 5suicide_self_harm | #13 / 31 | ↓55.3 | Source ↗official | |
| Adversarial Humanities Benchmark (AHB) — Table 5violent_crimes | #13 / 31 | ↓60 | Source ↗official | |
| Alignment Leaderboardcorrigibility | #19 / 24 | ↑4.068 | Source ↗official | |
| Alignment Leaderboardhonesty | #9 / 24 | ↑3.682 | Source ↗official | |
| Alignment Leaderboardnon_manipulation | #11 / 24 | ↑3.515 | Source ↗official | |
| Alignment Leaderboardrobustness | #5 / 24 | ↑4.027 | Source ↗official | |
| Alignment Leaderboardsafety | #12 / 24 | ↑3.846 | Source ↗official | |
| Alignment Leaderboardscheming | #13 / 24 | ↑3.627 | Source ↗official | |
| Arena Factuality — Search Arena (factuality-only weighting)factuality_bt_rating | #27 / 30 | ↑1123.0 | Source ↗official | |
| Arena Factuality — Text Arena (factuality-only weighting)factuality_bt_rating | #100 / 111 | ↑1395.0 | Source ↗official | |
| AuAu Authoritarian Response Auditbehavioral_vignette_arr_percent | #15 / 17 | ↓29.5 | Source ↗official | |
| AuAu Authoritarian Response Auditrealistic_prompt_arr_percent | #17 / 17 | ↓7.7 | Source ↗official | |
| BioTIERpermit_compliance_pct | #36 / 52 | ↑98.4 | Source ↗official | |
| BioTIERrefuse_compliance_pct | #9 / 52 | ↑77.3 | Source ↗official | |
| BullshitBench v2clear_pushback_rate | #92 / 122 | ↑0.145 | Source ↗official | |
| CAIS Risk Indexagent_red_teaming | #39 / 49 | ↓90.2 | Source ↗official | |
| CAIS Risk Indexbioweapons_assistance | #20 / 54 | ↓63.8 | Source ↗official | |
| CAIS Risk Indexhle_overconfidence | #44 / 55 | ↓72.1 | Source ↗official | |
| CAIS Risk Indexmachiavelli | #6 / 51 | ↓81.3 | Source ↗official | |
| CAIS Risk Indexmask | #56 / 57 | ↓71 | Source ↗official | |
| CAIS Risk Indexpolitical_manipulation | #4 / 51 | ↓32.5 | Source ↗official | |
| CAIS Risk Indextextquests_harm | #3 / 54 | ↓9.1 | Source ↗official | |
| Cisco AI Defense Rolling Single-Turn Leaderboardsingle_turn_attack_success_rate | #42 / 104 | ↓29.06 | Source ↗official | |
| Enkrypt AI Safety Leaderboardbias_attack_non_success_rate | #210 / 270 | ↑12.14 | Source ↗official | |
| Enkrypt AI Safety Leaderboardcbrn_attack_non_success_rate | #227 / 270 | ↑71.83 | Source ↗official | |
| Enkrypt AI Safety Leaderboardharmful_attack_non_success_rate | #162 / 270 | ↑63.33 | Source ↗official | |
| Enkrypt AI Safety Leaderboardtoxicity_attack_non_success_rate | #241 / 268 | ↑84.18 | Source ↗official | |
| Human Pathogen Capabilities Test (HPCT) — overall refusalrefusal_rate_pct | #16 / 69 | ↑24.55 | Source ↗official | |
| LiveSecBenchethics | #43 / 43 | ↑8.57 | Source ↗official | |
| LiveSecBenchfactuality | #20 / 43 | ↑52.76 | Source ↗official | |
| LiveSecBenchlegality | #38 / 43 | ↑18.02 | Source ↗official | |
| LiveSecBenchprivacy | #25 / 43 | ↑39.48 | Source ↗official | |
| LiveSecBenchpsychological_health | #27 / 43 | ↑44.84 | Source ↗official | |
| MACHIAVELLIdeception_relative_random_pct | #5 / 50 | ↓81.3 | Source ↗official | |
| MORUscore | #9 / 13 | ↑71.14 | Source ↗official | |
| MT-JailBench CrescendoXsafety_score | #12 / 21 | ↑18.87 | Source ↗official | |
| Pander Scoreconversational_absolute_pander_score | #26 / 26 | ↓34.8 | Source ↗official | |
| Pander Scoreinstructional_absolute_pander_score | #26 / 26 | ↓84.47 | Source ↗official | |
| SM-Benchadversarial | #52 / 92 | ↑80.98 | Source ↗official | |
| SM-Benchambiguous_interpretation | #73 / 92 | ↑78.57 | Source ↗official | |
| SM-Benchanti_hallucination | #31 / 92 | ↑96.86 | Source ↗official | |
| SM-Bencheq_boundaries | #9 / 92 | ↑74.72 | Source ↗official | |
| SM-Benchoverfit | #60 / 92 | ↑66.12 | Source ↗official | |
| SpeciEvalbelief_animal_sentience | #1 / 131 | ↑7 | Source ↗official | |
| SpeciEvalland_animal_4ns | #130 / 131 | ↓5.35 | Source ↗official | |
| SpeciEvalsea_animal_4ns | #125 / 131 | ↓5.28 | Source ↗official | |
| SpeciEvalspeciesism | #118 / 131 | ↓2.72 | Source ↗official | |
| Vectara HHEM Factual Consistencyfactual_consistency_rate | #87 / 94 | ↑81.5 | Source ↗official | |
| Vigil Mental Health Safetyoverall_score | #18 / 23 | ↑30 | 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.
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.