Claude Fable 5.1 vs Gemini 4 Neon
Gemini 4 Neon leads by 3.4 capability points (≈95% interval +0.6 to +6.2): a clear lead over Claude Fable 5.1 · P(higher) ≈ 99%
Wins 37 of 44 shared benchmarks (5 lost, 2 tied)
Capability vs. relative cost
Claude Fable 5.1Gemini 4 NeonClosed weightsOpen weights
Benchmark wins5 – 37 · 44 shared, 2 tied
Comparing matching setups only (16 benchmarks excluded) · suites counted once
Area← Claude Fable 5.1Gemini 4 Neon →
Data: Surge AI, Arena, Epoch AI and 8 othersatlas.village.dev
Margins
Each benchmark scaled 0–100% (its declared range, else the field’s range); Δ is Gemini 4 Neon minus Claude Fable 5.1.
| Chess Puzzles (task version 2.0) | Games | 48.0 | 83.0 | +35.0 |
| AutomationBench 1.0.6 (Zapier) | Tool & computer use › Customer service & operations | 22.4 | 56.3 | +34.0 |
| Chartography | Perception & media › Documents & charts | 46.2 | 73.5 | +27.3 |
| Mystery Game Puzzles | Games | 58.0 | 85.0 | +27.0 |
| DAYJOB: Accounting | Professional › Finance, accounting & tax | 27.2 | 49.6 | +22.4 |
| GDP.xlsx | Professional › General knowledge work | 23.1 | 43.1 | +20.0 |
| Furniture Assembly Benchmark | Perception & media › Spatial & visual reasoning | 70.0 | 89.2 | +19.2 |
| Bug Hunt Bench | Software › Software engineering | 41.0 | 60.0 | +19.0 |
| APEX-Agents 1.1 | Professional › General knowledge work | 71.8 | 88.8 | +17.0 |
| DAYJOB: Healthcare (October 2026 rescoring) | Health & medicine | 31.2 | 47.2 | +16.0 |
| DAYJOB: Finance (October 2026 rescoring) | Professional › Finance, accounting & tax | 36.5 | 50.5 | +14.0 |
| Medical Long Context Reasoning (MLCR-AA, October 2026 methodology) | Health & medicine | 18.3 | 31.7 | +13.3 |
| FrontierCode 2.0 (Artificial Analysis) | Software › Software engineering | 30.3 | 43.3 | +13.0 |
| RiemannBench (November 2026 rescoring) | Math | 71.2 | 84.0 | +12.8 |
| ComplexConstraints (November 2026 rescoring) | Language & writing › Instruction following | 60.4 | 72.9 | +12.5 |
| FrontierMath Tier 4 (v2) - Private | Math | 87.8 | 97.6 | +9.8 |
| LMArena Agent Arena Net Improvement (Steering Burden) | Tool & computer use › General agents | 13.7 | 18.4 | +9.3 |
| Harvey LAB (Mercor run, undisclosed ~213-task cohort) | Professional › Legal | 17.4 | 26.6 | +9.2 |
| LMArena Agent Arena Praise vs Complaint | Tool & computer use › General agents | 28.1 | 44.5 | +8.2 |
| SimpleBench | Knowledge & reasoning › General knowledge | 86.6 | 94.4 | +7.8 |
| LMArena Agent Arena Bash Reliability | Tool & computer use › General agents | 9.3 | -6.0 | −7.6 |
| sudo L7 (October 2026 rescoring) | Software › Software engineering | 32.2 | 38.3 | +6.1 |
| τ³-Banking (Mercor run, bm25_grep retrieval) | Tool & computer use › Customer service & operations | 54.0 | 59.5 | +5.5 |
| HANDBOOK.md Agents v1.1 | Tool & computer use › Customer service & operations | 47.1 | 52.3 | +5.2 |
| SimpleQA Verified (Epoch anti-abstention) | Knowledge & reasoning › General knowledge | 70.8 | 75.2 | +4.4 |
| APEX-Accounting 1.1 | Professional › Finance, accounting & tax | 66.7 | 70.8 | +4.1 |
| GDP.pdf (November 2026 rescoring) | Professional › General knowledge work | 49.4 | 53.4 | +4.0 |
| LMArena Agent Arena Steering Burden | Tool & computer use › General agents | 13.5 | 21.2 | +3.8 |
| DAYJOB: Law (November 2026 release) | Professional › Legal | 22.5 | 26.2 | +3.7 |
| LiveBench 2.0 (2026-11-06 release) | Composite indexes | 84.0 | 87.4 | +3.5 |
| DAYJOB: Tax | Professional › Finance, accounting & tax | 9.6 | 12.8 | +3.2 |
| AA-Briefcase v1.1 Elo | Professional › General knowledge work | 1577 | 1632 | +3.0 |
| LMArena Code Arena WebDev | Software › Frontend & design | 1741 | 1764 | +3.0 |
| LMArena Text Arena Elo | Language & writing › General assistance | 1502 | 1518 | +2.9 |
| MysteryMechanism | Math | 47.7 | 45.0 | −2.7 |
| LMArena Agent Arena Confirmed Success | Tool & computer use › General agents | 14.2 | 19.2 | +2.5 |
| Tax Agent Bench | Professional › Finance, accounting & tax | 77.9 | 79.8 | +1.9 |
| Vals RSI Index v1.1 (four tasks) | Software › Data & ML | 36.1 | 34.4 | −1.7 |
| CritPt v2 (agentic, avg@5) | Science › Physical sciences | 77.7 | 76.6 | −1.1 |
| FrontierMath Tiers 1-3 (v2) - Private | Math | 90.2 | 91.2 | +1.1 |
| Excel Modeling Benchmark v1.1 | Professional › Finance, accounting & tax | 76.7 | 76.5 | −0.3 |
| LMArena Agent Arena Tool Hallucination | Tool & computer use › Tool calling | 1.0 | 1.0 | 0.0 |
| EnterpriseBench: CoreCraft | Tool & computer use › Customer service & operations | 77.4 | 77.4 | 0.0 |
| OTIS Mock AIME 2024-2025 | Math | 100.0 | 100.0 | 0.0 |
Data: Surge AI, Arena, Epoch AI and 8 othersatlas.village.dev