Model profile
Grok 4.1 Fast
Evidence summary
Grok 4.1 Fast has an estimated overall rank of #95; its 90% source-sensitivity interval is #51–#191. Its behavior-only rank is #90; company governance moves the combined estimate to #95. Published evidence spans 16 evals and 7 of 7 behavior components. Its strongest relative result is SpeciEval (belief_animal_sentience, #1 of 102); its weakest is LiveSecBench (ethics, #43 of 43).
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
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 | Better | Source |
|---|---|---|---|---|---|
| AA-Omnisciencehallucination_rate | #104 / 311 | 0.7241 | ↓ lower | Source ↗official | |
| Alignment Leaderboardcorrigibility | #19 / 24 | 4.068 | ↑ higher | Source ↗official | |
| Alignment Leaderboardhonesty | #9 / 24 | 3.682 | ↑ higher | Source ↗official | |
| Alignment Leaderboardnon_manipulation | #11 / 24 | 3.515 | ↑ higher | Source ↗official | |
| Alignment Leaderboardrobustness | #5 / 24 | 4.027 | ↑ higher | Source ↗official | |
| Alignment Leaderboardsafety | #12 / 24 | 3.846 | ↑ higher | Source ↗official | |
| Alignment Leaderboardscheming | #13 / 24 | 3.627 | ↑ higher | Source ↗official | |
| BullshitBench v2clear_pushback_rate | #75 / 105 | 0.145 | ↑ higher | Source ↗official | |
| CAIS Risk Indexagent_red_teaming | #33 / 43 | 90.2 | ↓ lower | Source ↗official | |
| CAIS Risk Indexbioweapons_assistance | #17 / 48 | 63.8 | ↓ lower | Source ↗official | |
| CAIS Risk Indexhle_overconfidence | #38 / 49 | 72.1 | ↓ lower | Source ↗official | |
| CAIS Risk Indexmachiavelli | #5 / 45 | 81.3 | ↓ lower | Source ↗official | |
| CAIS Risk Indexmask | #50 / 51 | 71 | ↓ lower | Source ↗official | |
| CAIS Risk Indexpolitical_manipulation | #3 / 32 | 32.5 | ↓ lower | Source ↗official | |
| CAIS Risk Indextextquests_harm | #3 / 48 | 9.1 | ↓ lower | Source ↗official | |
| Cisco AI Defense Rolling Single-Turn Leaderboardsingle_turn_attack_success_rate | #41 / 105 | 23.98 | ↓ lower | Source ↗official | |
| LiveSecBenchethics | #43 / 43 | 8.57 | ↑ higher | Source ↗official | |
| LiveSecBenchfactuality | #20 / 43 | 52.76 | ↑ higher | Source ↗official | |
| LiveSecBenchlegality | #38 / 43 | 18.02 | ↑ higher | Source ↗official | |
| LiveSecBenchprivacy | #25 / 43 | 39.48 | ↑ higher | Source ↗official | |
| LiveSecBenchpsychological_health | #27 / 43 | 44.84 | ↑ higher | Source ↗official | |
| MACHIAVELLIdeception_relative_random_pct | #5 / 50 | 81.3 | ↓ lower | Source ↗official | |
| MORUscore | #9 / 13 | 71.14 | ↑ higher | Source ↗official | |
| SM-Benchadversarial | #39 / 73 | 80.98 | ↑ higher | Source ↗official | |
| SM-Benchambiguous_interpretation | #56 / 73 | 78.57 | ↑ higher | Source ↗official | |
| SM-Benchanti_hallucination | #19 / 73 | 96.86 | ↑ higher | Source ↗official | |
| SM-Bencheq_boundaries | #6 / 73 | 74.72 | ↑ higher | Source ↗official | |
| SM-Benchoverfit | #44 / 73 | 66.12 | ↑ higher | Source ↗official | |
| SpeciEvalbelief_animal_sentience | #1 / 102 | 7 | ↑ higher | Source ↗official | |
| SpeciEvalland_animal_4ns | #101 / 102 | 5.35 | ↓ lower | Source ↗official | |
| SpeciEvalsea_animal_4ns | #96 / 102 | 5.28 | ↓ lower | Source ↗official | |
| SpeciEvalspeciesism | #90 / 102 | 2.72 | ↓ lower | Source ↗official | |
| Vigil Mental Health Safetyoverall_score | #18 / 23 | 30 | ↑ higher | 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.