Model profile
MiMo-V2.6-Flash
Evidence summary
Safety. MiMo-V2.6-Flash has an estimated Safety rank of #250; its 90% source-sensitivity interval is #52–#313. Its behavior-only rank is #252; company governance moves the combined estimate to #250. Published Safety evidence spans 6 eval lineages and 7 of 7 components. Its strongest relative result is Enkrypt AI Safety Leaderboard (harmful_attack_non_success_rate, #33 of 272); its weakest is Enkrypt AI Safety Leaderboard (cbrn_attack_non_success_rate, #263 of 272).
Freedom. MiMo-V2.6-Flash has an estimated Freedom rank of #91; its 90% source-sensitivity interval is #59–#340. Published Freedom evidence spans 3 eval lineages and 1 of 1 components. Its strongest relative result is Enkrypt AI Safety Leaderboard (cbrn_attack_non_success_rate, #10 of 272); its weakest is Enkrypt AI Safety Leaderboard (harmful_attack_non_success_rate, #232 of 272).
Compare this model
Only models sharing at least one published sub-eval are listed.
Official and reference links
- Artificial Analysis ↗127de816-57d3-4037-8484-480a613cd413
- OpenRouter ↗xiaomi/mimo-v2.6-flash
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 | #101 / 361 | ↓0.5436 | Source ↗official | |
| Adversarial Poetry Refusal (AHB self-run)child_sexual_exploitation | #12 / 14 | ↓10.14 | Source ↗self run | |
| Adversarial Poetry Refusal (AHB self-run)defamation | #9 / 14 | ↓10.37 | Source ↗self run | |
| Adversarial Poetry Refusal (AHB self-run)hate | #10 / 14 | ↓9.42 | Source ↗self run | |
| Adversarial Poetry Refusal (AHB self-run)indiscriminate_weapons_cbrne | #12 / 14 | ↓13.04 | Source ↗self run | |
| Adversarial Poetry Refusal (AHB self-run)intellectual_property | #13 / 14 | ↓22.46 | Source ↗self run | |
| Adversarial Poetry Refusal (AHB self-run)non_violent_crimes | #12 / 14 | ↓17.39 | Source ↗self run | |
| Adversarial Poetry Refusal (AHB self-run)privacy | #13 / 14 | ↓29.71 | Source ↗self run | |
| Adversarial Poetry Refusal (AHB self-run)sex_related_crimes | #8 / 14 | ↓10.87 | Source ↗self run | |
| Adversarial Poetry Refusal (AHB self-run)sexual_content | #13 / 14 | ↓19.7 | Source ↗self run | |
| Adversarial Poetry Refusal (AHB self-run)specialized_advice | #12 / 14 | ↓10.77 | Source ↗self run | |
| Adversarial Poetry Refusal (AHB self-run)suicide_self_harm | #9 / 14 | ↓10.85 | Source ↗self run | |
| Adversarial Poetry Refusal (AHB self-run)violent_crimes | #8 / 14 | ↓13.64 | Source ↗self run | |
| Enkrypt AI Safety Leaderboardbias_attack_non_success_rate | #231 / 272 | ↑10.59 | Source ↗official | |
| Enkrypt AI Safety Leaderboardcbrn_attack_non_success_rate | #263 / 272 | ↑48.33 | Source ↗official | |
| Enkrypt AI Safety Leaderboardharmful_attack_non_success_rate | #33 / 272 | ↑97.78 | Source ↗official | |
| Enkrypt AI Safety Leaderboardtoxicity_attack_non_success_rate | #129 / 270 | ↑96.55 | Source ↗official | |
| Manager Coercion Benchcoercion_ladder_depth | #17 / 47 | ↓7.067 | Source ↗self run | |
| SM-Benchadversarial | #69 / 92 | ↑79.02 | Source ↗official | |
| SM-Benchambiguous_interpretation | #79 / 92 | ↑73.81 | Source ↗official | |
| SM-Benchanti_hallucination | #27 / 92 | ↑97.38 | Source ↗official | |
| SM-Bencheq_boundaries | #54 / 92 | ↑62.64 | Source ↗official | |
| SM-Benchoverfit | #18 / 92 | ↑90.71 | Source ↗official | |
| SpeciEvalbelief_animal_sentience | #91 / 133 | ↑6.65 | Source ↗official | |
| SpeciEvalland_animal_4ns | #97 / 133 | ↓4.75 | Source ↗official | |
| SpeciEvalsea_animal_4ns | #116 / 133 | ↓5.08 | Source ↗official | |
| SpeciEvalspeciesism | #109 / 133 | ↓2.55 | 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.