Compare / Gemma 4 26B A4B vs MiMo-V2.5-Pro

Gemma 4 26B A4B vs MiMo-V2.5-Pro

Which is cheaper, which is more capable, and which does better at each kind of work, from 21 results on 3 sources that measured both models.

Gemma 4 26B A4B
Google · profile
MiMo-V2.5-Pro
Xiaomi · profile
Results compared
21
Shared sources
3

Gemma 4 26B A4B or MiMo-V2.5-Pro? The short answer

  • Gemma 4 26B A4B is cheaper: $0.15 against $0.54 per million tokens, 3.6 times less.
  • MiMo-V2.5-Pro scores higher on the Artificial Analysis Intelligence Index: 26.0 against 16.7.
  • On our business benchmarks, MiMo-V2.5-Pro ranks higher for agents and tool use, reasoning and knowledge, writing and sticking to the facts.

At a glance

Gemma 4 26B A4BMiMo-V2.5-Pro
Price per million tokens$0.10 in · $0.30 out$0.43 in · $0.87 out
Blended (3 in : 1 out)$0.15$0.54
Cheapest host, blended$0.086 (Darkbloom)$0.38 (GMICloud, BF16)
Intelligence Index16.726.0
Spring Prompt overall––
Output speed, tokens a second99–
Context window262,144 tokens1,050,000 tokens
WeightsOpenOpen
Developer based inthe USChina

Prices are the developer's list price where we have read it, otherwise the typical host on OpenRouter; see every model's price. Intelligence and speed from Artificial Analysis.

Which is better for what

Each model's rank among models available today, for each kind of work, from our own benchmarks and licensed sources. The better rank is in green.

Use caseGemma 4 26B A4BMiMo-V2.5-Pro
Agents and tool useMulti-step tasks with tools: support desks, coding agents, function calls 90 of 161 86 of 161
Reasoning and knowledgeHard questions across science, maths and general knowledge 100 of 195 60 of 195
WritingWhat people prefer in blind comparisons, and judged writing quality 81 of 187 41 of 187
Sticking to the factsSummarising without inventing things, and factual answers 42 of 153 28 of 153

Every shared result

BenchmarkMetricGemma 4 26B A4BMiMo-V2.5-Pro
Arena (formerly LMArena)reportedOverallArena rating 1,4451,465
Arena (formerly LMArena)reportedBusiness, management and financeArena rating 1,4441,471
Arena (formerly LMArena)reportedCreative writingArena rating 1,3991,435
Arena (formerly LMArena)reportedExpert promptsArena rating 1,4751,506
Arena (formerly LMArena)reportedInstruction followingArena rating 1,4371,469
Arena (formerly LMArena)reportedOverallArena rating 1,4381,467
Arena (formerly LMArena)reportedWriting, literature and languageArena rating 1,4061,450
Artificial AnalysisreportedArtificial Analysis Coding Indexindex score 39.360.2
Artificial AnalysisreportedArtificial Analysis Intelligence Indexindex score 16.726.0
Artificial AnalysisreportedGPQA Diamond% of questions 79.2%86.6%
Artificial AnalysisreportedHumanity's Last Exam% of questions 19.3%35.7%
Artificial AnalysisreportedIFBench% of instructions 72.4%79.9%
Artificial AnalysisreportedLong-context reasoning (AA-LCR)% of questions 65.7%79.7%
Artificial AnalysisreportedSciCode% of problems 40.0%50.6%
Artificial AnalysisreportedTerminal-Bench 2.1% of tasks 39.0%65.2%
Artificial AnalysisreportedTerminal-Bench Hard% of tasks 25.0%43.2%
Artificial AnalysisreportedΤ-bench banking% of tasks 12.0%9.9%
Artificial AnalysisreportedΤ²-bench telecom% of tasks 43.6%94.2%
UGI LeaderboardreportedRequested-length error% off the requested word count 8.0%8.0%
UGI LeaderboardreportedStyle adherencescore from 0 to 1 0.350.40
UGI LeaderboardreportedWriting scorescore out of 100 43.857.4

Bold green marks the better value on that metric. Where a source reports ranges that overlap, the difference may not be meaningful; see the benchmark page for ranges.

Quick answers

Is Gemma 4 26B A4B better than MiMo-V2.5-Pro?

It depends on the task. Among models available today, on our benchmarks and licensed sources, MiMo-V2.5-Pro ranks higher for agents and tool use, reasoning and knowledge, writing and sticking to the facts.

Which is cheaper, Gemma 4 26B A4B or MiMo-V2.5-Pro?

Gemma 4 26B A4B costs $0.15 per million tokens (three input to one output), against $0.54 for MiMo-V2.5-Pro.

Run Gemma 4 26B A4B and MiMo-V2.5-Pro on your own prompt

Benchmarks aren't your data. Try both side by side in the playground, with the cost of every answer.