Gemini 4 Neon vs Muse Spark 1.3
Gemini 4 Neon leads by 7.9 capability points (≈95% interval +5.3 to +10.6): a clear lead over Muse Spark 1.3 · P(higher) ≈ 100%
Wins 53 of 55 shared benchmarks (2 lost)
Capability vs. relative cost
Gemini 4 NeonMuse Spark 1.3Closed weightsOpen weights
Benchmark wins53 – 2 · 55 shared
Comparing matching setups only (4 benchmarks excluded) · suites counted once
Area← Gemini 4 NeonMuse Spark 1.3 →
Data: Surge AI, Artificial Analysis, Arena and 10 othersatlas.village.dev
Margins
Each benchmark scaled 0–100% (its declared range, else the field’s range); Δ is Muse Spark 1.3 minus Gemini 4 Neon.
| Terminal-Bench-Science 0.1 | Science › Multidisciplinary | 70.0 | 3.8 | −66.2 |
| Mystery Game Puzzles | Games | 85.0 | 25.0 | −60.0 |
| Terminal-Bench Science 0.1 (Vals) | Science › Multidisciplinary | 62.9 | 4.3 | −58.6 |
| FrontierMath Tier 4 (v2) - Private | Math | 97.6 | 46.3 | −51.2 |
| Terminal-Bench 4.0 | Software › Software engineering | 61.6 | 10.6 | −51.0 |
| Terminal-Bench 4.0 (Artificial Analysis) | Software › Software engineering | 63.6 | 16.7 | −47.0 |
| Chess Puzzles (task version 2.0) | Games | 83.0 | 37.0 | −46.0 |
| Chartography | Perception & media › Documents & charts | 73.5 | 27.6 | −45.9 |
| Humanity's Sixth Sense | Perception & media › Spatial & visual reasoning | 75.9 | 37.4 | −38.5 |
| Humanity's Last Exam Diamond | Knowledge & reasoning › General knowledge | 71.3 | 33.7 | −37.6 |
| AutomationBench 1.0.6 (Zapier) | Tool & computer use › Customer service & operations | 56.3 | 20.7 | −35.6 |
| LMArena Agent Arena Net Improvement (Steering Burden) | Tool & computer use › General agents | 18.4 | 1.1 | −33.7 |
| DAYJOB: Accounting | Professional › Finance, accounting & tax | 49.6 | 18.0 | −31.6 |
| RiemannBench (November 2026 rescoring) | Math | 84.0 | 53.6 | −30.4 |
| Bug Hunt Bench | Software › Software engineering | 60.0 | 30.7 | −29.3 |
| sudo L7 (October 2026 rescoring) | Software › Software engineering | 38.3 | 12.8 | −25.5 |
| DAYJOB: Healthcare (October 2026 rescoring) | Health & medicine | 47.2 | 22.0 | −25.2 |
| Artificial Analysis Intelligence Index v5.0.1 | Composite indexes | 61.4 | 36.6 | −24.8 |
| CursorBench 4.1 | Software › Software engineering | 65.4 | 41.9 | −23.5 |
| DAYJOB: Finance (October 2026 rescoring) | Professional › Finance, accounting & tax | 50.5 | 27.0 | −23.5 |
| LMArena Agent Arena Praise vs Complaint | Tool & computer use › General agents | 44.5 | 0.1 | −22.2 |
| Artificial Analysis Coding Agent Index v2.1 | Software › Software engineering | 58.6 | 36.8 | −21.9 |
| AutomationBench-AA 1.0.6 | Tool & computer use › Customer service & operations | 79.3 | 57.9 | −21.5 |
| FrontierCode 2.0 (Artificial Analysis) | Software › Software engineering | 44.4 | 23.1 | −21.3 |
| GDP.xlsx | Professional › General knowledge work | 43.1 | 22.6 | −20.5 |
| LiveBench 2.0 (2026-11-06 release) | Composite indexes | 87.4 | 68.6 | −18.8 |
| APEX-Agents 1.1 | Professional › General knowledge work | 88.9 | 70.4 | −18.5 |
| HANDBOOK.md Agents v1.1 | Tool & computer use › Customer service & operations | 52.3 | 33.9 | −18.4 |
| FrontierMath Tiers 1-3 (v2) - Private | Math | 91.2 | 74.4 | −16.8 |
| Medical Long Context Reasoning (MLCR-AA, October 2026 methodology) | Health & medicine | 31.7 | 15.6 | −16.1 |
| EnterpriseBench: CoreCraft | Tool & computer use › Customer service & operations | 77.4 | 61.9 | −15.5 |
| AA-Omniscience v1.1 Index | Knowledge & reasoning › General knowledge | 54.8 | 23.8 | −15.5 |
| Vals RSI Index v1.1 (four tasks) | Software › Data & ML | 34.4 | 19.6 | −14.8 |
| LMArena Code Arena WebDev | Software › Frontend & design | 1764 | 1654 | −14.6 |
| DAYJOB: Law (November 2026 release) | Professional › Legal | 26.2 | 11.8 | −14.4 |
| SimpleBench | Knowledge & reasoning › General knowledge | 94.4 | 81.8 | −12.6 |
| AA-Briefcase v1.1 Elo | Professional › General knowledge work | 1699 | 1480 | −12.0 |
| Harvey LAB (Mercor run, undisclosed ~213-task cohort) | Professional › Legal | 26.6 | 15.0 | −11.6 |
| Vibe Code Bench v1.1 | Software › Frontend & design | 93.7 | 82.9 | −10.8 |
| LMArena Agent Arena Steering Burden | Tool & computer use › General agents | 21.2 | 0.8 | −10.2 |
| CritPt v2 (agentic, avg@5) | Science › Physical sciences | 78.0 | 68.0 | −10.0 |
| DAYJOB: Tax | Professional › Finance, accounting & tax | 12.8 | 3.6 | −9.2 |
| MysteryMechanism | Math | 45.0 | 36.0 | −9.0 |
| LMArena Agent Arena Confirmed Success | Tool & computer use › General agents | 19.2 | 2.5 | −8.4 |
| ComplexConstraints (November 2026 rescoring) | Language & writing › Instruction following | 72.9 | 64.8 | −8.1 |
| τ³-Banking (Mercor run, bm25_grep retrieval) | Tool & computer use › Customer service & operations | 59.5 | 52.2 | −7.3 |
| GDP.pdf (Artificial Analysis, All-pass) | Professional › General knowledge work | 33.4 | 26.6 | −6.8 |
| Tax Agent Bench | Professional › Finance, accounting & tax | 79.8 | 73.4 | −6.4 |
| APEX-Accounting 1.1 | Professional › Finance, accounting & tax | 70.8 | 64.9 | −5.9 |
| LMArena Text Arena Elo | Language & writing › General assistance | 1518 | 1498 | −3.6 |
| LMArena Agent Arena Bash Reliability | Tool & computer use › General agents | -6.0 | 1.2 | +3.6 |
| Legal Research Bench | Professional › Legal | 56.3 | 55.3 | −1.0 |
| OTIS Mock AIME 2024-2025 | Math | 100.0 | 99.2 | −0.8 |
| GDP.pdf (November 2026 rescoring) | Professional › General knowledge work | 53.4 | 53.0 | −0.4 |
| LMArena Agent Arena Tool Hallucination | Tool & computer use › Tool calling | 1.0 | 1.0 | 0.0 |
Data: Surge AI, Artificial Analysis, Arena and 10 othersatlas.village.dev