Side-by-side benchmark comparison across agentic, coding, multimodal, knowledge, reasoning, and math workflows.
Gemma 4 E2B
~39
Winner · 3/8 categoriesGPT-OSS 20B
36
1/8 categoriesGemma 4 E2B· GPT-OSS 20B
Pick Gemma 4 E2B if you want the stronger benchmark profile. GPT-OSS 20B only becomes the better choice if reasoning is the priority or you would rather avoid the extra latency and token burn of a reasoning model.
Gemma 4 E2B has the cleaner overall profile here, landing at 39 versus 36. It is a real lead, but still close enough that category-level strengths matter more than the headline number.
Gemma 4 E2B's sharpest advantage is in coding, where it averages 44 against 14.4. The single biggest benchmark swing on the page is BBH, 21.9% to 62%. GPT-OSS 20B does hit back in reasoning, so the answer changes if that is the part of the workload you care about most.
Gemma 4 E2B is the reasoning model in the pair, while GPT-OSS 20B is not. That usually helps on harder chain-of-thought-heavy tests, but it can also mean more latency and more token spend in real use.
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 E2B | GPT-OSS 20B |
|---|---|---|
| Agentic | ||
| Terminal-Bench 2.0 | — | 35% |
| BrowseComp | — | 42% |
| OSWorld-Verified | — | 31% |
| CodingGemma 4 E2B wins | ||
| LiveCodeBench | 44% | 11% |
| SWE-bench Verified | — | 14% |
| SWE-bench Pro | — | 18% |
| React Native Evals | — | 64.3% |
| Multimodal & GroundedGemma 4 E2B wins | ||
| MMMU-Pro | 44.2% | 31% |
| OfficeQA Pro | — | 42% |
| ReasoningGPT-OSS 20B wins | ||
| BBH | 21.9% | 62% |
| MRCRv2 | 19.1% | 48% |
| MuSR | — | 27% |
| LongBench v2 | — | 48% |
| KnowledgeGemma 4 E2B wins | ||
| GPQA | 43.4% | 30% |
| MMLU-Pro | 60% | — |
| MMLU | — | 85.3% |
| SuperGPQA | — | 28% |
| FrontierScience | — | 34% |
| SimpleQA | — | 29% |
| Instruction Following | ||
| IFEval | — | 67% |
| Multilingual | ||
| MGSM | — | 61% |
| MMLU-ProX | — | 59% |
| Mathematics | ||
| AIME 2023 | — | 31% |
| AIME 2024 | — | 33% |
| AIME 2025 | — | 32% |
| HMMT Feb 2025 | — | 28% |
| BRUMO 2025 | — | 30% |
| MATH-500 | — | 59% |
Gemma 4 E2B is ahead overall, 39 to 36. The biggest single separator in this matchup is BBH, where the scores are 21.9% and 62%.
Gemma 4 E2B has the edge for knowledge tasks in this comparison, averaging 54.1 versus 30.6. 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 14.4. Inside this category, LiveCodeBench is the benchmark that creates the most daylight between them.
GPT-OSS 20B has the edge for reasoning in this comparison, averaging 42.4 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 36. 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.