Side-by-side benchmark comparison across agentic, coding, multimodal, knowledge, reasoning, and math workflows.
Gemma 4 31B
73
Winner · 4/8 categoriesGPT-5.4 nano
58
0/8 categoriesGemma 4 31B· GPT-5.4 nano
Pick Gemma 4 31B if you want the stronger benchmark profile. GPT-5.4 nano only becomes the better choice if you need the larger 400K context window.
Gemma 4 31B is clearly ahead on the aggregate, 73 to 58. The gap is large enough that you do not need to squint at the spreadsheet to see the difference.
Gemma 4 31B's sharpest advantage is in reasoning, where it averages 66.4 against 38.7. The single biggest benchmark swing on the page is MRCRv2, 66.4% to 38.7%.
GPT-5.4 nano is also the more expensive model on tokens at $0.20 input / $1.25 output per 1M tokens, versus $0.00 input / $0.00 output per 1M tokens for Gemma 4 31B. That is roughly Infinityx on output cost alone. GPT-5.4 nano gives you the larger context window at 400K, compared with 256K for Gemma 4 31B.
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 31B | GPT-5.4 nano |
|---|---|---|
| Agentic | ||
| Terminal-Bench 2.0 | — | 46.3% |
| OSWorld-Verified | — | 39% |
| MCP Atlas | — | 56.1% |
| Toolathlon | — | 35.5% |
| Tau2-Telecom | — | 92.5% |
| CodingGemma 4 31B wins | ||
| LiveCodeBench | 80% | — |
| SWE-bench Pro | — | 52.4% |
| Multimodal & GroundedGemma 4 31B wins | ||
| MMMU-Pro | 76.9% | 66.1% |
| MMMU-Pro w/ Python | — | 69.5% |
| ReasoningGemma 4 31B wins | ||
| BBH | 74.4% | — |
| MRCRv2 | 66.4% | 38.7% |
| MRCR v2 64K-128K | — | 44.2% |
| MRCR v2 128K-256K | — | 33.1% |
| Graphwalks BFS 128K | — | 73.4% |
| Graphwalks Parents 128K | — | 50.8% |
| KnowledgeGemma 4 31B wins | ||
| GPQA | 84.3% | 82.8% |
| MMLU-Pro | 85.2% | — |
| HLE | 26.5% | 37.7% |
| HLE w/o tools | 19.5% | 24.3% |
| Instruction Following | ||
| Coming soon | ||
| Multilingual | ||
| Coming soon | ||
| Mathematics | ||
| Coming soon | ||
Gemma 4 31B is ahead overall, 73 to 58. The biggest single separator in this matchup is MRCRv2, where the scores are 66.4% and 38.7%.
Gemma 4 31B has the edge for knowledge tasks in this comparison, averaging 61.3 versus 53.2. Inside this category, HLE is the benchmark that creates the most daylight between them.
Gemma 4 31B has the edge for coding in this comparison, averaging 80 versus 52.4. GPT-5.4 nano stays close enough that the answer can still flip depending on your workload.
Gemma 4 31B has the edge for reasoning in this comparison, averaging 66.4 versus 38.7. Inside this category, MRCRv2 is the benchmark that creates the most daylight between them.
Gemma 4 31B has the edge for multimodal and grounded tasks in this comparison, averaging 76.9 versus 66.1. 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.