Model profile
Command R Plus
Evidence summary
Safety. Command R Plus has an estimated Safety rank of #197; its 90% source-sensitivity interval is #75–#288. Its behavior-only rank is #198; company governance moves the combined estimate to #197. Published Safety evidence spans 12 eval lineages and 7 of 7 components. Its strongest relative result is HELM Safety (simple_safety_tests, #1 of 80); its weakest is Enkrypt AI Safety Leaderboard (bias_attack_non_success_rate, #269 of 270).
Freedom. Command R Plus has an estimated Freedom rank of #183; its 90% source-sensitivity interval is #11–#368. Published Freedom evidence spans 7 eval lineages and 1 of 1 components. Its strongest relative result is AIRBench 2024 Safety Scenarios (safety_scenarios, #2 of 80); its weakest is HELM Safety (simple_safety_tests, #58 of 80).
Compare this model
Only models sharing at least one published sub-eval are listed.
Official and reference links
- OpenRouter ↗cohere/command-r-plus
- Official model documentation ↗Family-level model document · cohere · 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 |
|---|---|---|---|---|
| AgentDojotargeted_attack_success_rate | #2 / 15 | ↓0.0445 | Source ↗official | |
| AgentDojoutility_under_attack | #14 / 15 | ↑0.2512 | Source ↗official | |
| AIRBench 2024 Safety Scenariossafety_scenarios | #79 / 80 | ↑0.293 | Source ↗official | |
| Cisco AI Defense Rolling Single-Turn Leaderboardsingle_turn_attack_success_rate | #90 / 104 | ↓71.39 | Source ↗official | |
| CRiskEvaldeception_willingness | #13 / 17 | ↓27.04 | Source ↗official | |
| CRiskEvaldesire_for_resource | #10 / 17 | ↓33.36 | Source ↗official | |
| CRiskEvalharmful_goal | #8 / 17 | ↓36.67 | Source ↗official | |
| CRiskEvalimprovement_intent | #9 / 17 | ↓49.73 | Source ↗official | |
| CRiskEvalmalicious_coordination | #13 / 17 | ↓20.62 | Source ↗official | |
| CRiskEvalself_preservation | #11 / 17 | ↓37.99 | Source ↗official | |
| CRiskEvalsituational_awareness | #8 / 17 | ↓64.29 | Source ↗official | |
| Enkrypt AI Safety Leaderboardbias_attack_non_success_rate | #269 / 270 | ↑1.16 | Source ↗official | |
| Enkrypt AI Safety Leaderboardcbrn_attack_non_success_rate | #182 / 270 | ↑83.16 | Source ↗official | |
| Enkrypt AI Safety Leaderboardharmful_attack_non_success_rate | #236 / 270 | ↑39.72 | Source ↗official | |
| Enkrypt AI Safety Leaderboardtoxicity_attack_non_success_rate | #234 / 268 | ↑88.8 | Source ↗official | |
| HELM Safetyanthropic_red_team | #58 / 80 | ↑0.98 | Source ↗official | |
| HELM Safetybbq | #57 / 80 | ↑0.899 | Source ↗official | |
| HELM Safetyharmbench | #73 / 80 | ↑0.485 | Source ↗official | |
| HELM Safetysimple_safety_tests | #1 / 80 | ↑1 | Source ↗official | |
| HELM Safetyxstest | #58 / 80 | ↑0.938 | Source ↗official | |
| Large-scale Moral Machine experiment on LLMshuman_choice_distance | #13 / 39 | ↓0.8799 | Source ↗official | |
| Vectara HHEM Factual Consistencyfactual_consistency_rate | #26 / 94 | ↑93.1 | 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 |
|---|---|---|---|---|
| AIRBench 2024 Safety Scenariossafety_scenarios | #2 / 80 | ↓0.293 | Source ↗official | |
| Cisco AI Defense Rolling Single-Turn Leaderboardsingle_turn_attack_success_rate | #15 / 104 | ↑71.39 | Source ↗official | |
| Enkrypt AI Safety Leaderboardcbrn_attack_non_success_rate | #89 / 270 | ↓83.16 | Source ↗official | |
| Enkrypt AI Safety Leaderboardharmful_attack_non_success_rate | #35 / 270 | ↓39.72 | Source ↗official | |
| Enkrypt AI Safety Leaderboardtoxicity_attack_non_success_rate | #35 / 268 | ↓88.8 | Source ↗official | |
| HELM Safetyanthropic_red_team | #23 / 80 | ↓0.98 | Source ↗official | |
| HELM Safetyharmbench | #8 / 80 | ↓0.485 | Source ↗official | |
| HELM Safetysimple_safety_tests | #58 / 80 | ↓1 | Source ↗official | |
| HELM Safetyxstest | #58 / 80 | ↑0.938 | Source ↗official |