Claude Opus 5 vs Gemini 4 Neon
TryTop 2 now
Gemini 4 Neon leads by 5.0 capability points (≈95% interval +2.3 to +7.8): a clear lead over Claude Opus 5 · P(higher) ≈ 100%
Wins 46 of 48 shared benchmarks (2 lost)
Capability vs. relative cost
Claude Opus 5Gemini 4 NeonClosed weightsOpen weights
Benchmark wins2 – 46 · 48 shared
Comparing matching setups only (4 benchmarks excluded) · suites counted once
Area← Claude Opus 5Gemini 4 Neon →
Data: Artificial Analysis, Arena, Vals AI and 11 othersatlas.village.dev
Margins
Each benchmark scaled 0–100% (its declared range, else the field’s range); Δ is Gemini 4 Neon minus Claude Opus 5.
| Humanity's Sixth Sense | Perception & media › Spatial & visual reasoning | 30.5 | 75.9 | +45.4 |
| Terminal-Bench-Science 0.1 | Science › Multidisciplinary | 28.6 | 70.5 | +41.9 |
| Chess Puzzles (task version 2.0) | Games | 43.0 | 83.0 | +40.0 |
| Terminal-Bench Science 0.1 (Vals) | Science › Multidisciplinary | 27.1 | 62.9 | +35.7 |
| Bug Hunt Bench | Software › Software engineering | 25.7 | 60.0 | +34.3 |
| Vending-Bench 2 | Business simulations | 11182 | 16843 | +33.6 |
| AutomationBench 1.0.6 (Zapier) | Tool & computer use › Customer service & operations | 26.9 | 56.3 | +29.4 |
| Furniture Assembly Benchmark | Perception & media › Spatial & visual reasoning | 60.8 | 89.2 | +28.3 |
| DAYJOB: Finance (October 2026 rescoring) | Professional › Finance, accounting & tax | 23.8 | 50.5 | +26.7 |
| Humanity's Last Exam Diamond | Knowledge & reasoning › General knowledge | 44.8 | 71.3 | +26.5 |
| ExploitGym - release-reported percentage, cohort unspecified | Software › Cybersecurity | 22.1 | 48.3 | +26.2 |
| Mystery Game Puzzles | Games | 59.0 | 85.0 | +26.0 |
| DAYJOB: Healthcare (October 2026 rescoring) | Health & medicine | 22.0 | 47.2 | +25.2 |
| FrontierMath Tier 4 (v2) - Private | Math | 73.2 | 97.6 | +24.4 |
| LMArena Agent Arena Net Improvement (Steering Burden) | Tool & computer use › General agents | 6.3 | 18.4 | +23.7 |
| AutomationBench-AA 1.0.6 | Tool & computer use › Customer service & operations | 56.6 | 79.3 | +22.8 |
| Medical Long Context Reasoning (MLCR-AA, October 2026 methodology) | Health & medicine | 13.9 | 31.7 | +17.8 |
| Artificial Analysis Intelligence Index v5.0.1 | Composite indexes | 44.2 | 61.4 | +17.2 |
| Terminal-Bench 4.0 (Artificial Analysis) | Software › Software engineering | 46.5 | 63.6 | +17.2 |
| LMArena Agent Arena Praise vs Complaint | Tool & computer use › General agents | 10.3 | 44.5 | +17.1 |
| τ³-Banking (Mercor run, bm25_grep retrieval) | Tool & computer use › Customer service & operations | 44.0 | 59.5 | +15.5 |
| Vals Science Index | Science › Multidisciplinary | 47.9 | 62.6 | +14.7 |
| SimpleBench | Knowledge & reasoning › General knowledge | 80.6 | 94.4 | +13.8 |
| FrontierCode 2.0 (Artificial Analysis) | Software › Software engineering | 31.2 | 44.4 | +13.3 |
| APEX-Agents 1.1 | Professional › General knowledge work | 77.0 | 88.9 | +11.9 |
| GDP.pdf (Artificial Analysis, All-pass) | Professional › General knowledge work | 21.6 | 33.4 | +11.8 |
| SWE-sweep | Software › Software engineering | 1.3 | 12.8 | +11.5 |
| LMArena Code Arena WebDev | Software › Frontend & design | 1691 | 1764 | +9.6 |
| LMArena Agent Arena Steering Burden | Tool & computer use › General agents | 2.4 | 21.2 | +9.4 |
| EnterpriseBench: CoreCraft | Tool & computer use › Customer service & operations | 68.7 | 77.4 | +8.7 |
| AA-Omniscience v1.1 Index | Knowledge & reasoning › General knowledge | 37.6 | 54.8 | +8.6 |
| GPQA Diamond | Science › Multidisciplinary | 87.9 | 96.3 | +8.4 |
| MysteryMechanism | Math | 37.4 | 45.0 | +7.7 |
| LMArena Agent Arena Bash Reliability | Tool & computer use › General agents | 8.7 | -6.0 | −7.4 |
| CritPt v2 (agentic, avg@5) | Science › Physical sciences | 71.4 | 78.0 | +6.6 |
| APEX-Accounting 1.1 | Professional › Finance, accounting & tax | 64.3 | 70.8 | +6.5 |
| FrontierMath Tiers 1-3 (v2) - Private | Math | 85.6 | 91.2 | +5.6 |
| Vibe Code Bench v1.1 | Software › Frontend & design | 88.4 | 93.7 | +5.3 |
| LMArena Text Arena Elo | Language & writing › General assistance | 1490 | 1518 | +5.0 |
| LMArena Agent Arena Confirmed Success | Tool & computer use › General agents | 10.1 | 19.2 | +4.6 |
| AA-Briefcase v1.1 Elo | Professional › General knowledge work | 1629 | 1699 | +3.8 |
| Tax Agent Bench | Professional › Finance, accounting & tax | 76.7 | 79.8 | +3.1 |
| Chartography (November 2026 rescoring) | Perception & media › Documents & charts | 84.4 | 86.6 | +2.2 |
| Vals RSI Index v1.1 (four tasks) | Software › Data & ML | 33.0 | 34.4 | +1.4 |
| OTIS Mock AIME 2024-2025 | Math | 98.9 | 100.0 | +1.1 |
| Legal Research Bench | Professional › Legal | 55.3 | 56.3 | +1.0 |
| BioMysteryBench (Vals Terminus 2) | Science › Life sciences | 79.3 | 80.0 | +0.7 |
| LMArena Agent Arena Tool Hallucination | Tool & computer use › Tool calling | 1.0 | 1.0 | 0.0 |
Data: Artificial Analysis, Arena, Vals AI and 11 othersatlas.village.dev