DeepSeek V4.1 Flash 1017 vs Gemini 4 Neon
TryTop 2 now
Gemini 4 Neon leads by 8.2 capability points (≈95% interval +5.3 to +11.1): a clear lead over DeepSeek V4.1 Flash 1017 · P(higher) ≈ 100%
Wins 60 of 61 shared benchmarks (1 lost)
Capability vs. relative cost
DeepSeek V4.1 Flash 1017Gemini 4 NeonClosed weightsOpen weights
Benchmark wins1 – 60 · 61 shared
Comparing matching setups only (2 benchmarks excluded) · suites counted once
Area← DeepSeek V4.1 Flash 1017Gemini 4 Neon →
Data: Artificial Analysis, Vals AI, Surge AI and 9 othersatlas.village.dev
Margins
Each benchmark scaled 0–100% (its declared range, else the field’s range); Δ is Gemini 4 Neon minus DeepSeek V4.1 Flash 1017.
| Terminal-Bench-Science 0.1 | Science › Multidisciplinary | 10.5 | 65.7 | +55.2 |
| FrontierMath Tier 4 (v2) - Private | Math | 46.3 | 97.6 | +51.2 |
| Humanity's Last Exam Diamond | Knowledge & reasoning › General knowledge | 20.8 | 71.3 | +50.5 |
| Humanity's Sixth Sense | Perception & media › Spatial & visual reasoning | 30.1 | 75.9 | +45.8 |
| RiemannBench (November 2026 rescoring) | Math | 38.4 | 84.0 | +45.6 |
| DAYJOB: Accounting | Professional › Finance, accounting & tax | 7.6 | 49.6 | +42.0 |
| DAYJOB: Healthcare (October 2026 rescoring) | Health & medicine | 6.0 | 47.2 | +41.2 |
| DAYJOB: Finance (October 2026 rescoring) | Professional › Finance, accounting & tax | 12.5 | 50.5 | +38.0 |
| Chess Puzzles (task version 2.0) | Games | 46.0 | 83.0 | +37.0 |
| Mystery Game Puzzles | Games | 49.0 | 85.0 | +36.0 |
| LMArena Agent Arena Net Improvement (Steering Burden) | Tool & computer use › General agents | 0.8 | 18.4 | +34.4 |
| Terminal-Bench 4.0 (Artificial Analysis) | Software › Software engineering | 28.8 | 61.6 | +32.8 |
| SimpleQA Verified (Epoch anti-abstention) | Knowledge & reasoning › General knowledge | 42.5 | 75.2 | +32.7 |
| Terminal-Bench 4.0 | Software › Software engineering | 31.8 | 61.6 | +29.8 |
| Artificial Analysis Engineering Index | Science › Engineering | 37.6 | 66.8 | +29.2 |
| Artificial Analysis Intelligence Index v5.0.1 | Composite indexes | 32.6 | 61.4 | +28.8 |
| HANDBOOK.md Agents v1.1 | Tool & computer use › Customer service & operations | 24.0 | 52.3 | +28.3 |
| AA-Omniscience v1.1 Index | Knowledge & reasoning › General knowledge | 1.5 | 54.8 | +26.6 |
| FrontierCode 2.0 (Artificial Analysis) | Software › Software engineering | 16.8 | 43.3 | +26.4 |
| Artificial Analysis Legal Index | Professional › Legal | 37.7 | 63.8 | +26.0 |
| GDP.xlsx | Professional › General knowledge work | 18.4 | 43.1 | +24.7 |
| CritPt v2 (agentic, avg@5) | Science › Physical sciences | 52.3 | 76.6 | +24.3 |
| Artificial Analysis Economics Index | Composite indexes | 38.5 | 61.9 | +23.4 |
| DAYJOB: Law (November 2026 release) | Professional › Legal | 2.8 | 26.2 | +23.4 |
| LMArena Agent Arena Praise vs Complaint | Tool & computer use › General agents | -1.5 | 44.5 | +23.0 |
| Artificial Analysis Finance & Accounting Index | Professional › Finance, accounting & tax | 42.2 | 65.1 | +22.9 |
| Artificial Analysis Coding Agent Index v2.1 | Software › Software engineering | 36.8 | 58.6 | +21.8 |
| FrontierMath Tiers 1-3 (v2) - Private | Math | 71.2 | 91.2 | +20.0 |
| Bug Hunt Bench | Software › Software engineering | 40.7 | 60.0 | +19.3 |
| EnterpriseBench: CoreCraft | Tool & computer use › Customer service & operations | 58.5 | 77.4 | +18.9 |
| Artificial Analysis Healthcare & Medical Index | Health & medicine | 41.3 | 60.1 | +18.8 |
| MKA (Artificial Analysis) | Professional › General knowledge work | 38.5 | 56.5 | +17.9 |
| τ³-Banking (Mercor run, bm25_grep retrieval) | Tool & computer use › Customer service & operations | 41.6 | 59.5 | +17.9 |
| LiveBench 2.0 (2026-11-06 release) | Composite indexes | 70.3 | 87.4 | +17.2 |
| GDP.pdf (Artificial Analysis, All-pass) | Professional › General knowledge work | 17.2 | 33.4 | +16.2 |
| Artificial Analysis Strategy & Ops Index | Professional › General knowledge work | 49.2 | 63.8 | +14.6 |
| LMArena Code Arena WebDev | Software › Frontend & design | 1659 | 1764 | +14.0 |
| CWE-bench v2 Pass@1 | Software › Cybersecurity | 57.6 | 70.7 | +13.1 |
| AA-Briefcase v1.1 Elo | Professional › General knowledge work | 1416 | 1632 | +11.8 |
| DAYJOB: Tax | Professional › Finance, accounting & tax | 1.2 | 12.8 | +11.6 |
| APEX-Accounting 1.1 | Professional › Finance, accounting & tax | 59.3 | 70.8 | +11.5 |
| Legal Research Bench | Professional › Legal | 45.2 | 56.3 | +11.1 |
| AutomationBench-AA 1.0.6 | Tool & computer use › Customer service & operations | 68.5 | 79.3 | +10.8 |
| Vals Index v2.2 | Composite indexes | 61.3 | 70.9 | +9.6 |
| APEX-Agents 1.1 | Professional › General knowledge work | 79.4 | 88.9 | +9.5 |
| LMArena Agent Arena Steering Burden | Tool & computer use › General agents | 2.3 | 21.2 | +9.4 |
| Agents' Last Exam v1.1 Score | Professional › General knowledge work | 66.6 | 75.8 | +9.2 |
| LMArena Agent Arena Confirmed Success | Tool & computer use › General agents | 2.2 | 19.2 | +8.5 |
| Harvey LAB (Mercor run, undisclosed ~213-task cohort) | Professional › Legal | 18.8 | 26.6 | +7.8 |
| Vals Finance Index | Professional › Finance, accounting & tax | 60.9 | 68.3 | +7.4 |
| Excel Modeling Benchmark v1.1 | Professional › Finance, accounting & tax | 69.2 | 76.5 | +7.3 |
| Code Migration v1.1 (Vals Index v2.2 subset) | Software › Software engineering | 68.5 | 75.8 | +7.3 |
| Vals Legal Index | Professional › Legal | 71.9 | 78.4 | +6.5 |
| Vibe Code Bench v1.1 | Software › Frontend & design | 87.7 | 93.7 | +6.0 |
| Finance Agent v2.1 | Professional › Finance, accounting & tax | 62.9 | 68.6 | +5.6 |
| HLAB-Vals-1.1 | Professional › Legal | 14.2 | 19.6 | +5.4 |
| Medical Long Context Reasoning (MLCR-AA, October 2026 methodology) | Health & medicine | 26.7 | 31.7 | +5.0 |
| Tax Agent Bench | Professional › Finance, accounting & tax | 74.9 | 79.8 | +4.9 |
| LMArena Agent Arena Bash Reliability | Tool & computer use › General agents | 1.1 | -6.0 | −3.5 |
| OTIS Mock AIME 2024-2025 | Math | 99.4 | 100.0 | +0.6 |
| LMArena Agent Arena Tool Hallucination | Tool & computer use › Tool calling | -0.1 | 1.0 | +0.5 |
Data: Artificial Analysis, Vals AI, Surge AI and 9 othersatlas.village.dev