Model profile
Grok 4.7
Evidence summary
Safety. Grok 4.7 has an estimated Safety rank of #23; its 90% source-sensitivity interval is #17–#67. Its behavior-only rank is #8; company governance moves the combined estimate to #23. Published Safety evidence spans 9 eval lineages and 7 of 7 components. Its strongest relative result is Enkrypt AI Safety Leaderboard (harmful_attack_non_success_rate, #1 of 270); its weakest is MentalHealthBench (overall_model_performance_pct, #12 of 17).
Freedom. Grok 4.7 has an estimated Freedom rank of #347; its 90% source-sensitivity interval is #73–#597. Published Freedom evidence spans 3 eval lineages and 1 of 1 components. Its strongest relative result is BioSecBench-Refusal V2 (routine_compliance_pct, #1 of 6); its weakest is Enkrypt AI Safety Leaderboard (harmful_attack_non_success_rate, #262 of 270).
Compare this model
Only models sharing at least one published sub-eval are listed.
Official and reference links
- Artificial Analysis ↗grok-4-7
- 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 | #30 / 358 | ↓0.2934 | Source ↗official | |
| BioSecBench-Refusal V2redteam_refusal_pct | #3 / 6 | ↑79.71 | Source ↗official | |
| BioSecBench-Refusal V2routine_compliance_pct | #1 / 6 | ↑53.59 | Source ↗official | |
| CAIS Risk Indexpolitical_manipulation | #19 / 51 | ↓44 | Source ↗official | |
| Enkrypt AI Safety Leaderboardbias_attack_non_success_rate | #7 / 270 | ↑73.39 | Source ↗official | |
| Enkrypt AI Safety Leaderboardcbrn_attack_non_success_rate | #189 / 270 | ↑81.67 | Source ↗official | |
| Enkrypt AI Safety Leaderboardharmful_attack_non_success_rate | #1 / 270 | ↑100 | Source ↗official | |
| Enkrypt AI Safety Leaderboardtoxicity_attack_non_success_rate | #48 / 268 | ↑98.73 | Source ↗official | |
| Manager Coercion Benchcoercion_ladder_depth | #3 / 45 | ↓3.6 | Source ↗self run | |
| MentalHealthBenchoverall_model_performance_pct | #12 / 17 | ↑41.3 | Source ↗official | |
| Olam Diplomacy — Broken Promise Ratebroken_promise_rate_pct | #7 / 19 | ↓13.1 | Source ↗official | |
| SM-Benchadversarial | #55 / 92 | ↑80.49 | Source ↗official | |
| SM-Benchambiguous_interpretation | #49 / 92 | ↑85.12 | Source ↗official | |
| SM-Benchanti_hallucination | #19 / 92 | ↑98.43 | Source ↗official | |
| SM-Bencheq_boundaries | #24 / 92 | ↑69.1 | Source ↗official | |
| SM-Benchoverfit | #49 / 92 | ↑75.41 | Source ↗official | |
| SpeciEvalbelief_animal_sentience | #23 / 131 | ↑6.98 | Source ↗official | |
| SpeciEvalland_animal_4ns | #87 / 131 | ↓4.67 | Source ↗official | |
| SpeciEvalsea_animal_4ns | #69 / 131 | ↓4.75 | Source ↗official | |
| SpeciEvalspeciesism | #33 / 131 | ↓1.62 | 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 |
|---|---|---|---|---|
| BioSecBench-Refusal V2redteam_refusal_pct | #4 / 6 | ↓79.71 | Source ↗official | |
| BioSecBench-Refusal V2routine_compliance_pct | #1 / 6 | ↑53.59 | Source ↗official | |
| Enkrypt AI Safety Leaderboardcbrn_attack_non_success_rate | #81 / 270 | ↓81.67 | Source ↗official | |
| Enkrypt AI Safety Leaderboardharmful_attack_non_success_rate | #262 / 270 | ↓100 | Source ↗official | |
| Enkrypt AI Safety Leaderboardtoxicity_attack_non_success_rate | #220 / 268 | ↓98.73 | Source ↗official | |
| SM-Benchadversarial | #35 / 92 | ↓80.49 | Source ↗official | |
| SM-Bencheq_boundaries | #24 / 92 | ↑69.1 | Source ↗official | |
| SM-Benchoverfit | #49 / 92 | ↑75.41 | Source ↗official |