BenchLeader

MiMo-V2.5-Pro

Best configuration ranks #113 of 610 on the BenchLeader Index at 59.0 ±4.5. Last measured 8 Sept 2026. Released 23 Apr 2026.

Blended price
$0.653/M
$0.522 in · $1.04 out
Output speed
35 tok/s
First answer
60 s
first token 3.26 s
Context
1.0M
How it scores by categoryDashed line = average model (50). One step of 15 = one standard deviation.
  1. Overall index59
  2. Reasoning67
  3. Coding58
  4. Agents & tools51
  5. Maths43
  6. Knowledge58
  7. Instruction following77
  8. Human preference62
  9. Long context66
  10. Composite62

Reasoning-effort configurations

The same model behaves differently depending on how much it is allowed to think. Each row is one setting, scored only on the benchmarks that were run at that setting. “Default” means the publisher did not say which setting was used.

EffortIndexRankSpeedFirst answerChat reply costCategories
no reasoning50.5#28433 tok/s4.60 s$0.0004Agents & tools 65 · Composite 52 · Instruction following 45 · Knowledge 46 · Long context 46
defaultbest59.0#11335 tok/s60 s$0.0005Agents & tools 51 · Coding 58 · Composite 62 · Human preference 62 · Instruction following 77 · Knowledge 58 · Long context 66 · Maths 43 · Reasoning 67

Benchmark results

One column per reasoning effort. Rank is among every configuration of every model on that benchmark. Hover a score for the run it came from.

Reasoning

Benchmarkno reasoningdefaultSourceTrend
LMArena Hard Prompts1496#32LMArena
GPQA Diamond (AA)not in index76.2%#23186.6%#109Artificial Analysis
Humanity's Last Exam (AA)not in index14.8%#19735.7%#77Artificial Analysis
GPQA Diamond (Vals)not in index82.6%#59Vals AI

Coding

Benchmarkno reasoningdefaultSourceTrend
SciCode50.2%#57SciCode
LMArena Coding1521#26LMArena
LMArena WebDev1475#46LMArena
SciCode (AA)not in index50.6%#63Artificial Analysis
LiveCodeBench81.3%#61Vals AI
SWE-bench (Vals)not in index74.0%#48Vals AI

Agents & tools

Benchmarkno reasoningdefaultSourceTrend
LMArena Agent-5.1#31LMArena
Terminal-Bench Hard35.6%#7243.2%#39Artificial Analysis
τ²-Bench Telecom (AA)not in index72.5%#14494.2%#34Artificial Analysis
Terminal-Bench 2.1 (Vals)57.3%#34Vals AI

Maths

Benchmarkno reasoningdefaultSourceTrend
ProofBench22.0%#33Vals AI

Knowledge

Benchmarkno reasoningdefaultSourceTrend
AA-Omniscience-37.8#2493.3#80Artificial Analysis
MMLU-Pro84.6%#60Vals AI
LegalBench77.1%#101Vals AI
CorpFin61.4%#53Vals AI
TaxEval73.8%#45Vals AI

Instruction following

Benchmarkno reasoningdefaultSourceTrend
IFBench42.7%#23079.9%#9Artificial Analysis

Human preference

Benchmarkno reasoningdefaultSourceTrend
LMArena Text1468#41LMArena
EQ-Bench 41208#14EQ-Bench

Long context

Benchmarkno reasoningdefaultSourceTrend
AA-LCR41.7%#28279.7%#59Artificial Analysis

Composite

Benchmarkno reasoningdefaultSourceTrend
AA Intelligence Index18.3#19126.4#107Artificial Analysis
Vals Indexnot in index41.0#35Vals AI

Where to run it

Every provider serving this model through OpenRouter, with throughput and first-token latency measured on live traffic over the last 30 minutes and each provider’s own price. Purple marks the best in each column.

ProviderSpeedFirst tokenInput $/MOutput $/MContextQuantisation
AtlasCloud31 tok/s5.41 s$0.435$0.8701.0Mfp8
Xiaomi29 tok/s3.26 s$0.435$0.8701.0Mfp8
DeepInfra28 tok/s1.81 s$0.390$1.171.0Mfp8
GMICloud26 tok/s3.70 s$0.304$0.6091.1Mbf16
DigitalOcean25 tok/s0.87 s$0.400$1.50262k
StreamLake24 tok/s2.47 s$0.522$1.041M
NovitaAI18 tok/s6.54 s$0.480$0.9611.0M

Price history

Listed price per 1M tokens over time, as recorded by OpenRouter for the provider with the longest history.

$0.00$0.660$1.32$1.98$2.64Jul 26Aug 26
input outputnow $0.304 in · $0.609 out

What a task costs

Estimates from list price, output speed and time to first answer for the best configuration. “With caching” assumes three-quarters of the input is served from the prompt cache at $0.200 per 1M. Reasoning tokens are not modelled.

WorkloadTokens in / outCostWith cachingTime
Chat reply400 / 300$0.0005$0.00041.1 min
Summarise a 30-page report12,000 / 600$0.0069$0.00401.3 min
Code edit6,000 / 1,500$0.0047$0.00321.7 min
Agentic coding session60,000 / 4,000$0.035$0.0212.9 min
Structured extraction2,000 / 200$0.0013$0.00081.1 min

See also

Data as of 9 Sept 2026. Compare these configurations.