Gemini 4 Neon vs GPT-6 Astra
TryTop 2 now
Gemini 4 Neon leads by 1.8 capability points (≈95% interval −1.1 to +4.8): statistically tied · P(higher) ≈ 89%
Wins 49 of 62 shared benchmarks (11 lost, 2 tied)
Capability vs. relative cost
Gemini 4 NeonGPT-6 AstraClosed weightsOpen weights
Benchmark wins49 – 11 · 62 shared, 2 tied
Comparing matching setups only (40 benchmarks excluded) · suites counted once
Area← Gemini 4 NeonGPT-6 Astra →
Data: Vals AI, Surge AI, Artificial Analysis and 9 othersatlas.village.dev
Margins
Each benchmark scaled 0–100% (its declared range, else the field’s range); Δ is GPT-6 Astra minus Gemini 4 Neon.
| EnterpriseBench: CoreCraft | Tool & computer use › Customer service & operations | 77.4 | 42.6 | −34.8 |
| Humanity's Sixth Sense | Perception & media › Spatial & visual reasoning | 75.9 | 53.6 | −22.3 |
| HANDBOOK.md Agents v1.1 | Tool & computer use › Customer service & operations | 52.3 | 31.4 | −20.9 |
| GDP.xlsx | Professional › General knowledge work | 43.1 | 22.9 | −20.2 |
| DAYJOB: Accounting | Professional › Finance, accounting & tax | 49.6 | 30.0 | −19.6 |
| DAYJOB: Healthcare (October 2026 rescoring) | Health & medicine | 47.2 | 28.4 | −18.8 |
| Harvey LAB (Mercor run, undisclosed ~213-task cohort) | Professional › Legal | 26.6 | 9.2 | −17.4 |
| Legal Research Bench | Professional › Legal | 56.3 | 39.4 | −16.8 |
| Tax Agent Bench | Professional › Finance, accounting & tax | 79.8 | 63.3 | −16.5 |
| AutomationBench 1.0.6 (Zapier) | Tool & computer use › Customer service & operations | 56.3 | 41.4 | −14.9 |
| APEX-Agents 1.1 | Professional › General knowledge work | 88.9 | 75.1 | −13.8 |
| DAYJOB: Finance (October 2026 rescoring) | Professional › Finance, accounting & tax | 50.5 | 37.3 | −13.2 |
| τ³-Banking (Mercor run, bm25_grep retrieval) | Tool & computer use › Customer service & operations | 59.5 | 46.7 | −12.8 |
| Vals Finance Index | Professional › Finance, accounting & tax | 68.3 | 55.6 | −12.7 |
| HLAB-Vals-1.1 | Professional › Legal | 19.6 | 7.1 | −12.5 |
| AutomationBench-AA 1.0.6 | Tool & computer use › Customer service & operations | 79.3 | 68.5 | −10.8 |
| SimpleBench | Knowledge & reasoning › General knowledge | 94.4 | 83.6 | −10.8 |
| DAYJOB: Law (November 2026 release) | Professional › Legal | 26.2 | 36.6 | +10.4 |
| Medical Long Context Reasoning (MLCR-AA, October 2026 methodology) | Health & medicine | 31.7 | 41.7 | +10.0 |
| Vals Legal Index | Professional › Legal | 78.4 | 68.4 | −10.0 |
| Furniture Assembly Benchmark | Perception & media › Spatial & visual reasoning | 89.2 | 80.0 | −9.2 |
| Chess Puzzles (task version 2.0) | Games | 83.0 | 74.0 | −9.0 |
| AA-Briefcase v1.1 Elo | Professional › General knowledge work | 1699 | 1544 | −8.5 |
| Finance Agent v2.1 | Professional › Finance, accounting & tax | 68.6 | 60.4 | −8.1 |
| MysteryMechanism | Math | 45.0 | 53.2 | +8.1 |
| DAYJOB: Tax | Professional › Finance, accounting & tax | 12.8 | 4.8 | −8.0 |
| LMArena Text Arena Elo | Language & writing › General assistance | 1518 | 1475 | −7.6 |
| LMArena Agent Arena Net Improvement (Steering Burden) | Tool & computer use › General agents | 18.4 | 14.6 | −7.5 |
| Humanity's Last Exam Diamond | Knowledge & reasoning › General knowledge | 71.3 | 63.8 | −7.5 |
| sudo L7 (October 2026 rescoring) | Software › Software engineering | 38.3 | 31.1 | −7.2 |
| MKA (Artificial Analysis) | Professional › General knowledge work | 56.5 | 50.2 | −6.2 |
| LMArena Agent Arena Bash Reliability | Tool & computer use › General agents | -6.0 | 6.5 | +6.2 |
| Terminal-Bench-Science 0.1 | Science › Multidisciplinary | 65.7 | 59.5 | −6.2 |
| LMArena Agent Arena Praise vs Complaint | Tool & computer use › General agents | 44.5 | 32.3 | −6.1 |
| APEX-Accounting 1.1 | Professional › Finance, accounting & tax | 70.8 | 64.7 | −6.1 |
| CritPt v2 (agentic, avg@5) | Science › Physical sciences | 78.0 | 84.0 | +6.0 |
| AA-Omniscience v1.1 Index | Knowledge & reasoning › General knowledge | 54.8 | 44.6 | −5.1 |
| SEC-Bench Pro - release-reported cohort | Software › Cybersecurity | 90.5 | 85.4 | −5.1 |
| Artificial Analysis Intelligence Index v5.0.1 | Composite indexes | 61.4 | 56.9 | −4.5 |
| Terminal-Bench 4.0 (Artificial Analysis) | Software › Software engineering | 63.6 | 59.6 | −4.0 |
| Excel Modeling Benchmark v1.1 | Professional › Finance, accounting & tax | 76.5 | 72.7 | −3.8 |
| LMArena Agent Arena Steering Burden | Tool & computer use › General agents | 21.2 | 13.9 | −3.6 |
| LiveBench 2.0 (2026-11-06 release) | Composite indexes | 87.4 | 84.3 | −3.2 |
| FrontierCode 2.0 (Artificial Analysis) | Software › Software engineering | 44.4 | 41.4 | −3.0 |
| GDP.pdf (November 2026 rescoring) | Professional › General knowledge work | 53.4 | 50.4 | −3.0 |
| Chartography | Perception & media › Documents & charts | 73.5 | 71.0 | −2.5 |
| FrontierMath Tiers 1-3 (v2) - Private | Math | 91.2 | 93.7 | +2.5 |
| Vals Science Index | Science › Multidisciplinary | 62.6 | 65.1 | +2.5 |
| RiemannBench (November 2026 rescoring) | Math | 84.0 | 81.6 | −2.4 |
| Code Migration v1.1 (Vals Index v2.2 subset) | Software › Software engineering | 75.8 | 78.0 | +2.3 |
| LMArena Agent Arena Confirmed Success | Tool & computer use › General agents | 19.2 | 14.9 | −2.2 |
| ComplexConstraints (November 2026 rescoring) | Language & writing › Instruction following | 72.9 | 70.8 | −2.1 |
| Terminal-Bench 4.0 | Software › Software engineering | 61.6 | 59.6 | −2.0 |
| GDP.pdf (Artificial Analysis, All-pass) | Professional › General knowledge work | 33.4 | 32.2 | −1.2 |
| Mystery Game Puzzles | Games | 85.0 | 84.0 | −1.0 |
| BioMysteryBench (Vals Terminus 2) | Science › Life sciences | 80.0 | 79.3 | −0.7 |
| GPQA Diamond | Science › Multidisciplinary | 96.3 | 95.8 | −0.5 |
| SimpleQA Verified (Epoch anti-abstention) | Knowledge & reasoning › General knowledge | 75.2 | 75.6 | +0.4 |
| LMArena Agent Arena Tool Hallucination | Tool & computer use › Tool calling | 1.0 | 1.2 | +0.1 |
| FrontierMath Tier 4 (v2) - Private | Math | 97.6 | 97.6 | 0.0 |
| OTIS Mock AIME 2024-2025 | Math | 100.0 | 100.0 | 0.0 |
| Terminal-Bench Science 0.1 (Vals) | Science › Multidisciplinary | 62.9 | 62.9 | 0.0 |
Data: Vals AI, Surge AI, Artificial Analysis and 9 othersatlas.village.dev