Side-by-side benchmark comparison across agentic, coding, multimodal, knowledge, reasoning, and math workflows.
Claude Opus 4.5
76
Winner · 3/8 categoriesGemma 4 31B
73
1/8 categoriesClaude Opus 4.5· Gemma 4 31B
Pick Claude Opus 4.5 if you want the stronger benchmark profile. Gemma 4 31B only becomes the better choice if coding is the priority or you need the larger 256K context window.
Claude Opus 4.5 has the cleaner overall profile here, landing at 76 versus 73. It is a real lead, but still close enough that category-level strengths matter more than the headline number.
Claude Opus 4.5's sharpest advantage is in knowledge, where it averages 73.2 against 61.3. The single biggest benchmark swing on the page is LiveCodeBench, 57% to 80%. Gemma 4 31B does hit back in coding, so the answer changes if that is the part of the workload you care about most.
Gemma 4 31B is the reasoning model in the pair, while Claude Opus 4.5 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. Gemma 4 31B gives you the larger context window at 256K, compared with 200K for Claude Opus 4.5.
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 | Claude Opus 4.5 | Gemma 4 31B |
|---|---|---|
| Agentic | ||
| Terminal-Bench 2.0 | 59.3% | — |
| BrowseComp | 73% | — |
| OSWorld-Verified | 66.3% | — |
| OSWorld | 66.3% | — |
| Claw-Eval | 59.6% | — |
| QwenClawBench | 52.3% | — |
| QwenWebBench | 1518 | — |
| TAU3-Bench | 70.2% | — |
| VITA-Bench | 23.3% | — |
| DeepPlanning | 26.4% | — |
| Toolathlon | 43.5% | — |
| MCP Atlas | 42.3% | — |
| MCP-Tasks | 71.8% | — |
| WideResearch | 76.4% | — |
| CodingGemma 4 31B wins | ||
| HumanEval | 91% | — |
| SWE-bench Verified | 80.9% | — |
| LiveCodeBench | 57% | 80% |
| LiveCodeBench v6 | 84.8% | — |
| SWE-bench Pro | 57.1% | — |
| SWE Multilingual | 77.5% | — |
| NL2Repo | 43.2% | — |
| Multimodal & GroundedClaude Opus 4.5 wins | ||
| MMMU | 80.7% | — |
| MMMU-Pro | 70.6% | 76.9% |
| OfficeQA Pro | 87% | — |
| RealWorldQA | 77.0% | — |
| OmniDocBench 1.5 | 87.7% | — |
| Video-MME (with subtitle) | 77.6% | — |
| Video-MME (w/o subtitle) | 81.4% | — |
| MathVision | 74.3% | — |
| We-Math | 70.0% | — |
| DynaMath | 79.7% | — |
| MStar | 73.2% | — |
| SimpleVQA | 65.7% | — |
| ChatCVQA | 68.5% | — |
| CountBench | 90.6% | — |
| AI2D_TEST | 87.7% | — |
| ERQA | 46.8% | — |
| VideoMMMU | 84.4% | — |
| MLVU (M-Avg) | 81.7% | — |
| ScreenSpot Pro | 45.7% | — |
| ReasoningClaude Opus 4.5 wins | ||
| MuSR | 93% | — |
| BBH | 87% | 74.4% |
| LongBench v2 | 64.4% | — |
| MRCRv2 | 81% | 66.4% |
| AI-Needle | 74% | — |
| ARC-AGI-2 | 37.6% | — |
| KnowledgeClaude Opus 4.5 wins | ||
| MMLU | 90.8% | — |
| GPQA | 87% | 84.3% |
| SuperGPQA | 70.6% | — |
| MMLU-Pro | 89.5% | 85.2% |
| MMLU-Redux | 96.6% | — |
| C-Eval | 92.2% | — |
| HLE | 30.8% | 26.5% |
| FrontierScience | 84% | — |
| SimpleQA | 95% | — |
| HLE w/o tools | — | 19.5% |
| Instruction Following | ||
| IFEval | 90.9% | — |
| IFBench | 58.0% | — |
| Multilingual | ||
| MGSM | 90% | — |
| MMLU-ProX | 85.7% | — |
| NOVA-63 | 56.7% | — |
| INCLUDE | 86.2% | — |
| PolyMath | 79.0% | — |
| VWT2k-lite | 79.7% | — |
| MAXIFE | 79.2% | — |
| Mathematics | ||
| AIME 2023 | 99% | — |
| AIME 2024 | 99% | — |
| AIME 2025 | 98% | — |
| AIME26 | 95.1% | — |
| HMMT Feb 2023 | 95% | — |
| HMMT Feb 2024 | 97% | — |
| HMMT Feb 2025 | 96% | — |
| HMMT Feb 2025 | 92.9% | — |
| HMMT Nov 2025 | 93.3% | — |
| HMMT Feb 2026 | 85.3% | — |
| MMAnswerBench | 84.0% | — |
| BRUMO 2025 | 96% | — |
| MATH-500 | 89% | — |
Claude Opus 4.5 is ahead overall, 76 to 73. The biggest single separator in this matchup is LiveCodeBench, where the scores are 57% and 80%.
Claude Opus 4.5 has the edge for knowledge tasks in this comparison, averaging 73.2 versus 61.3. 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 62.6. Inside this category, LiveCodeBench is the benchmark that creates the most daylight between them.
Claude Opus 4.5 has the edge for reasoning in this comparison, averaging 67.6 versus 66.4. Inside this category, MRCRv2 is the benchmark that creates the most daylight between them.
Claude Opus 4.5 has the edge for multimodal and grounded tasks in this comparison, averaging 78 versus 76.9. 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.