Side-by-side benchmark comparison across agentic, coding, multimodal, knowledge, reasoning, and math workflows.
DeepSeek V3.1 (Reasoning)
43
Winner · 1/8 categoriesGemma 4 E2B
~39
3/8 categoriesDeepSeek V3.1 (Reasoning)· Gemma 4 E2B
Pick DeepSeek V3.1 (Reasoning) if you want the stronger benchmark profile. Gemma 4 E2B only becomes the better choice if coding is the priority.
DeepSeek V3.1 (Reasoning) is clearly ahead on the aggregate, 43 to 39. The gap is large enough that you do not need to squint at the spreadsheet to see the difference.
DeepSeek V3.1 (Reasoning)'s sharpest advantage is in reasoning, where it averages 49.5 against 19.1. The single biggest benchmark swing on the page is BBH, 64% to 21.9%. Gemma 4 E2B 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 | DeepSeek V3.1 (Reasoning) | Gemma 4 E2B |
|---|---|---|
| Agentic | ||
| Terminal-Bench 2.0 | 42% | — |
| BrowseComp | 48% | — |
| CodingGemma 4 E2B wins | ||
| HumanEval | 26% | — |
| SWE-bench Verified | 14% | — |
| LiveCodeBench | 16% | 44% |
| SWE-bench Pro | 25% | — |
| Multimodal & GroundedGemma 4 E2B wins | ||
| MMMU-Pro | 37% | 44.2% |
| OfficeQA Pro | 47% | — |
| ReasoningDeepSeek V3.1 (Reasoning) wins | ||
| MuSR | 30% | — |
| BBH | 64% | 21.9% |
| LongBench v2 | 57% | — |
| MRCRv2 | 56% | 19.1% |
| KnowledgeGemma 4 E2B wins | ||
| MMLU | 34% | — |
| GPQA | 33% | 43.4% |
| SuperGPQA | 31% | — |
| MMLU-Pro | 53% | 60% |
| HLE | 10% | — |
| FrontierScience | 37% | — |
| SimpleQA | 32% | — |
| Instruction Following | ||
| IFEval | 70% | — |
| Multilingual | ||
| MGSM | 64% | — |
| MMLU-ProX | 61% | — |
| Mathematics | ||
| AIME 2023 | 34% | — |
| AIME 2024 | 36% | — |
| AIME 2025 | 35% | — |
| HMMT Feb 2025 | 31% | — |
| BRUMO 2025 | 33% | — |
| MATH-500 | 62% | — |
DeepSeek V3.1 (Reasoning) is ahead overall, 43 to 39. The biggest single separator in this matchup is BBH, where the scores are 64% and 21.9%.
Gemma 4 E2B has the edge for knowledge tasks in this comparison, averaging 54.1 versus 32.5. Inside this category, GPQA is the benchmark that creates the most daylight between them.
Gemma 4 E2B has the edge for coding in this comparison, averaging 44 versus 19. Inside this category, LiveCodeBench is the benchmark that creates the most daylight between them.
DeepSeek V3.1 (Reasoning) has the edge for reasoning in this comparison, averaging 49.5 versus 19.1. Inside this category, BBH is the benchmark that creates the most daylight between them.
Gemma 4 E2B has the edge for multimodal and grounded tasks in this comparison, averaging 44.2 versus 41.5. 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.