Side-by-side benchmark comparison across agentic, coding, multimodal, knowledge, reasoning, and math workflows.
Gemma 4 26B A4B
64
1/8 categoriesMiMo-V2-Flash
67
Winner · 3/8 categoriesGemma 4 26B A4B· MiMo-V2-Flash
Pick MiMo-V2-Flash if you want the stronger benchmark profile. Gemma 4 26B A4B only becomes the better choice if coding is the priority.
MiMo-V2-Flash has the cleaner overall profile here, landing at 67 versus 64. It is a real lead, but still close enough that category-level strengths matter more than the headline number.
MiMo-V2-Flash's sharpest advantage is in reasoning, where it averages 68.3 against 44.1. The single biggest benchmark swing on the page is MRCRv2, 44.1% to 73%. Gemma 4 26B A4B does hit back in coding, so the answer changes if that is the part of the workload you care about most.
BenchLM keeps the benchmark table and the operator tradeoffs on the same page so a better score does not hide a materially slower, pricier, or smaller-context model.
Runtime metrics show N/A when BenchLM does not have a sourced snapshot for that exact model. The scoring rules and freshness policy are documented on the methodology page.
| Benchmark | Gemma 4 26B A4B | MiMo-V2-Flash |
|---|---|---|
| Agentic | ||
| Terminal-Bench 2.0 | — | 63% |
| BrowseComp | — | 65% |
| OSWorld-Verified | — | 58% |
| Claw-Eval | — | 48.1% |
| CodingGemma 4 26B A4B wins | ||
| LiveCodeBench | 77.1% | 80.6% |
| HumanEval | — | 84.8% |
| SWE-bench Verified | — | 73.4% |
| SWE-bench Pro | — | 52% |
| SWE Multilingual | — | 71.7% |
| Multimodal & GroundedMiMo-V2-Flash wins | ||
| MMMU-Pro | 73.8% | 78% |
| OfficeQA Pro | — | 73% |
| ReasoningMiMo-V2-Flash wins | ||
| BBH | 64.8% | 85% |
| MRCRv2 | 44.1% | 73% |
| MuSR | — | 74% |
| LongBench v2 | — | 60.6% |
| KnowledgeMiMo-V2-Flash wins | ||
| GPQA | 82.3% | 83.7% |
| MMLU-Pro | 82.6% | 84.9% |
| HLE | 17.2% | 14% |
| HLE w/o tools | 8.7% | — |
| MMLU | — | 86.7% |
| SuperGPQA | — | 76% |
| FrontierScience | — | 71% |
| SimpleQA | — | 76% |
| Instruction Following | ||
| IFEval | — | 84% |
| Multilingual | ||
| MMLU-ProX | — | 77% |
| Mathematics | ||
| AIME 2023 | — | 79% |
| AIME 2024 | — | 81% |
| AIME 2025 | — | 94.1% |
| HMMT Feb 2023 | — | 75% |
| HMMT Feb 2024 | — | 77% |
| HMMT Feb 2025 | — | 76% |
| BRUMO 2025 | — | 78% |
| MATH-500 | — | 90% |
MiMo-V2-Flash is ahead overall, 67 to 64. The biggest single separator in this matchup is MRCRv2, where the scores are 44.1% and 73%.
MiMo-V2-Flash has the edge for knowledge tasks in this comparison, averaging 63.7 versus 56.1. Inside this category, HLE is the benchmark that creates the most daylight between them.
Gemma 4 26B A4B has the edge for coding in this comparison, averaging 77.1 versus 67.9. Inside this category, LiveCodeBench is the benchmark that creates the most daylight between them.
MiMo-V2-Flash has the edge for reasoning in this comparison, averaging 68.3 versus 44.1. Inside this category, MRCRv2 is the benchmark that creates the most daylight between them.
MiMo-V2-Flash has the edge for multimodal and grounded tasks in this comparison, averaging 75.8 versus 73.8. Inside this category, MMMU-Pro is the benchmark that creates the most daylight between them.
Get notified when new models drop, benchmark scores change, or the leaderboard shifts. One email per week.
Free. No spam. Unsubscribe anytime. We only store derived location metadata for consent routing.