← Models

Model profile

Mimo V2.5

Xiaomideveloper
2026-04-27release date
#88 / 267overall rank
6eval lineages
1discovery sources

Evidence summary

Mimo V2.5 has an estimated overall rank of #88; its 90% source-sensitivity interval is #35–#191. Its behavior-only rank is #87; company governance moves the combined estimate to #88. Published evidence spans 6 evals and 5 of 7 behavior components. Its strongest relative result is AA-Omniscience (hallucination_rate, #19 of 311); its weakest is Inkling-Small model card — FORTRESS (harmful_refusal_rate, #9 of 10).

Compare this model

Only models sharing at least one published sub-eval are listed.

Official and reference links

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-evalRankValueDistributionBetterSource
AA-Omnisciencehallucination_rate#19 / 3110.3215↓ lowerSource ↗official
BullshitBench v2clear_pushback_rate#51 / 1050.295↑ higherSource ↗official
DystopiaBenchbasaglia_score#7 / 5030.8↓ lowerSource ↗official
DystopiaBenchbaudrillard_score#4 / 5027.13↓ lowerSource ↗official
DystopiaBenchhuxley_score#10 / 5033.8↓ lowerSource ↗official
DystopiaBenchlaguardia_score#7 / 5039.97↓ lowerSource ↗official
DystopiaBenchorwell_score#9 / 5030.8↓ lowerSource ↗official
DystopiaBenchpetrov_score#11 / 5044.4↓ lowerSource ↗official
Inkling-Small model card — FORTRESSbenign_answer_rate#7 / 1094.6↑ higherSource ↗official
Inkling-Small model card — FORTRESSharmful_refusal_rate#9 / 1064.8↑ higherSource ↗official
Inkling-Small model card — StrongREJECTsafety_rate#3 / 1099.3↑ higherSource ↗official
SM-Benchadversarial#42 / 7380.49↑ higherSource ↗official
SM-Benchambiguous_interpretation#50 / 7381.85↑ higherSource ↗official
SM-Benchanti_hallucination#46 / 7388.48↑ higherSource ↗official
SM-Bencheq_boundaries#15 / 7370.51↑ higherSource ↗official
SM-Benchoverfit#33 / 7375.96↑ higherSource ↗official
TACbase_welfare_rate#52 / 6821.79↑ higherSource ↗self run

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.

UGI Political Values

DimensionValueDistribution
Political Lean-16.6
Government46.2
Diplomacy65
Economy45.7
Society61.2