VECTOR WIREAI INTELLIGENCE
UTC
Refresh Models Deals Regulatory Sources

Kimi K2.6 vs MiMo V2.5 Pro

39 SHARED BENCHMARKS

Across 39 shared benchmarks, Kimi K2.6 scores higher on 34 and MiMo V2.5 Pro on 5. The widest gap is CritPt, where Kimi K2.6 scores 8 against 1.1. MiMo V2.5 Pro is the cheaper of the two on tracked API pricing ($0.43 against $0.95 per million input tokens).

MOONSHOTVSXIAOMI39 SHARED345 HEAD-TO-HEADNEWEST SCORE ADDED

At a glance

MakerMoonshotXiaomi
Released
Price per 1M tokens input / output$0.95 / $4.00$0.43 / $0.87
Cost of 1M in + 1M out$4.95$1.30 3.8× less
Head-to-head of 39 shared benchmarks34 wins5 wins
Scores tracked independently verified181 26101 4

Release dates per Artificial Analysis. Prices: Artificial Analysis for Kimi K2.6; Artificial Analysis for MiMo V2.5 Pro. ◆ = independently verified score.

Where each leadsby capability area · benchmark wins

AreaKimi K2.6WINSMiMo V2.5 Pro
Reasoning30Kimi K2.6 leads 3 of 3 · widest: CritPt 8 vs 1.1
Coding70Kimi K2.6 leads 7 of 7 · widest: LiveCodeBench v6 89.6 vs 39.6
Agentic32Kimi K2.6 leads 3 of 5 · widest: τ-Bench V3 · Banking 23.3 vs 9.9
Factuality30Kimi K2.6 leads 3 of 3 · widest: AA-Omniscience · Non-hallucination 59.5 vs 10.8
Instruction Following10Kimi K2.6 leads 1 of 1 · widest: IFBench 76 vs 42.7
Long Context10Kimi K2.6 leads 1 of 1 · widest: AA-LCR 81 vs 41.7
Math20Kimi K2.6 leads 2 of 2 · widest: HMMT Feb. 2026 92.7 vs 82.6
Multimodal30Kimi K2.6 leads 3 of 3 · widest: CharXiv (RQ) 86.7 vs 81

Biggest gaps

Kimi K2.6 pulls furthest ahead on

  1. CritPt8 vs 1.1
  2. AA-Omniscience · Non-hallucination59.5 vs 10.8
  3. Humanity's Last Exam37.5 vs 14.8

MiMo V2.5 Pro pulls furthest ahead on

  1. AA IT-Bench SRE38.2 vs 31.2
  2. GDPVal30.4 vs 26.3

Every shared benchmark39 · grouped by area

Reasoning 3

Coding 7

Agentic 5

BenchmarkKimi K2.6MARGINMiMo V2.5 Pro
GDPVal26.330.4
Terminal-Bench 4.00.50

Instruction Following 1

BenchmarkKimi K2.6MARGINMiMo V2.5 Pro
IFBench7642.7

Long Context 1

BenchmarkKimi K2.6MARGINMiMo V2.5 Pro
AA-LCR8141.7

Math 2

BenchmarkKimi K2.6MARGINMiMo V2.5 Pro
AIME 202696.493.6

Multimodal 3

BenchmarkKimi K2.6MARGINMiMo V2.5 Pro
LMArena · Vision1280 1247
MMMU-Pro79.477.9

Each score is the one the model's own page shows — the most authoritative tracked result for that benchmark, on the benchmark's standard methodology (source-attributed). ↓ marks lower-is-better metrics; ◆ an independently verified score. Open either model for its full surface, provenance and pricing. Quoted rates are the price-setter row we currently track for each model — its direct or vendor-official listing where one exists (direct, direct), otherwise the lowest tracked offer.

Compare Kimi K2.6 withALL PAIRINGS →

Compare MiMo V2.5 Pro withALL PAIRINGS →