Atlas

Compare Models

Gemini 4 Neon vs Grok 4.7

Gemini 4 Neon leads by 8.4 capability points (≈95% interval +5.7 to +11.0): a clear lead over Grok 4.7 · P(higher) ≈ 100%

Wins 70 of 72 shared benchmarks (2 lost)

Capability vs. relative cost
Gemini 4 NeonGrok 4.7Closed weightsOpen weights

Benchmark wins70 – 2 · 72 shared

Comparing matching setups only (2 benchmarks excluded) · suites counted once
← Gemini 4 NeonGrok 4.7 →
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 Grok 4.7 minus Gemini 4 Neon.

FrontierMath Tier 4 (v2) - PrivateMath97.617.1−80.5
Furniture Assembly BenchmarkPerception & media › Spatial & visual reasoning89.220.8−68.3
ChartographyPerception & media › Documents & charts73.514.7−58.8
Terminal-Bench-Science 0.1Science › Multidisciplinary70.011.4−58.6
Mystery Game PuzzlesGames85.029.0−56.0
Terminal-Bench Science 0.1 (Vals)Science › Multidisciplinary62.910.0−52.9
Humanity's Last Exam DiamondKnowledge & reasoning › General knowledge71.323.6−47.7
Chess Puzzles (task version 2.0)Games83.041.0−42.0
FrontierMath Tiers 1-3 (v2) - PrivateMath91.253.0−38.2
Terminal-Bench 4.0 (Artificial Analysis)Software › Software engineering63.625.8−37.9
Vending-Bench 2Business simulations1684310537−37.4
LMArena Agent Arena Net Improvement (Steering Burden)Tool & computer use › General agents18.41.5−33.0
Terminal-Bench 4.0Software › Software engineering61.628.8−32.8
Bug Hunt BenchSoftware › Software engineering60.027.6−32.4
DAYJOB: AccountingProfessional › Finance, accounting & tax49.620.0−29.6
Vals Science IndexScience › Multidisciplinary62.634.8−27.8
τ³-Banking (Mercor run, bm25_grep retrieval)Tool & computer use › Customer service & operations59.532.6−26.9
RiemannBench (November 2026 rescoring)Math84.057.6−26.4
sudo L7 (October 2026 rescoring)Software › Software engineering38.312.2−26.1
Artificial Analysis Intelligence Index v5.0.1Composite indexes61.435.5−25.9
Artificial Analysis Engineering IndexScience › Engineering66.841.8−25.1
DAYJOB: Finance (October 2026 rescoring)Professional › Finance, accounting & tax50.526.5−24.0
Code Migration v1.1 (Vals Index v2.2 subset)Software › Software engineering75.851.8−23.9
LMArena Agent Arena Praise vs ComplaintTool & computer use › General agents44.52.1−21.2
GDP.xlsxProfessional › General knowledge work43.122.3−20.8
MysteryMechanismMath45.025.2−19.8
DAYJOB: Healthcare (October 2026 rescoring)Health & medicine47.227.6−19.6
LiveBench 2.0 (2026-11-06 release)Composite indexes87.467.9−19.5
SimpleQA Verified (Epoch anti-abstention)Knowledge & reasoning › General knowledge75.256.0−19.2
Artificial Analysis Coding Agent Index v2.1Software › Software engineering58.641.6−17.0
CursorBench 4.1Software › Software engineering65.448.6−16.8
LMArena Code Arena WebDevSoftware › Frontend & design17641639−16.6
Artificial Analysis Legal IndexProfessional › Legal63.847.2−16.6
MKA (Artificial Analysis)Professional › General knowledge work56.540.1−16.3
Artificial Analysis Economics IndexComposite indexes61.945.7−16.2
CWE-bench v2 Pass@1Software › Cybersecurity70.754.7−16.0
FrontierCode 2.0 (Artificial Analysis)Software › Software engineering44.428.7−15.8
CritPt v2 (agentic, avg@5)Science › Physical sciences78.062.3−15.7
APEX-Accounting 1.1Professional › Finance, accounting & tax70.555.3−15.2
Vals RSI Index v1.1 (four tasks)Software › Data & ML34.420.2−14.2
AutomationBench-AA 1.0.6Tool & computer use › Customer service & operations79.365.6−13.8
Artificial Analysis Healthcare & Medical IndexHealth & medicine60.146.6−13.5
Artificial Analysis Finance & Accounting IndexProfessional › Finance, accounting & tax65.151.7−13.4
LMArena Text Arena EloLanguage & writing › General assistance15181443−13.2
LMArena Agent Arena Steering BurdenTool & computer use › General agents21.2-4.3−12.7
Vals Index v2.2Composite indexes70.958.6−12.3
ComplexConstraints (November 2026 rescoring)Language & writing › Instruction following72.960.8−12.1
AA-Omniscience v1.1 IndexKnowledge & reasoning › General knowledge54.831.2−11.8
Tax Agent BenchProfessional › Finance, accounting & tax79.869.0−10.8
BioMysteryBench (Vals Terminus 2)Science › Life sciences80.069.3−10.7
Medical Long Context Reasoning (MLCR-AA, October 2026 methodology)Health & medicine31.721.1−10.6
GDP.pdf (Artificial Analysis, All-pass)Professional › General knowledge work33.423.2−10.2
Vals Finance IndexProfessional › Finance, accounting & tax68.358.8−9.5
EnterpriseBench: CoreCraftTool & computer use › Customer service & operations77.468.2−9.2
Legal Research BenchProfessional › Legal56.347.1−9.1
Artificial Analysis Strategy & Ops IndexProfessional › General knowledge work63.854.8−8.9
APEX-Agents 1.1Professional › General knowledge work88.980.4−8.5
HANDBOOK.md Agents v1.1Tool & computer use › Customer service & operations52.344.0−8.3
Finance Agent v2.1Professional › Finance, accounting & tax68.660.6−8.0
LMArena Agent Arena Confirmed SuccessTool & computer use › General agents19.23.5−7.8
DAYJOB: TaxProfessional › Finance, accounting & tax12.85.0−7.8
Vibe Code Bench v1.1Software › Frontend & design93.786.2−7.5
Harvey LAB (Mercor run, undisclosed ~213-task cohort)Professional › Legal26.619.5−7.1
LMArena Agent Arena Bash ReliabilityTool & computer use › General agents-6.08.2+7.1
Vals Legal IndexProfessional › Legal78.472.7−5.7
Excel Modeling Benchmark v1.1Professional › Finance, accounting & tax76.571.5−5.0
GDP.pdf (November 2026 rescoring)Professional › General knowledge work53.448.8−4.6
AA-Briefcase v1.1 EloProfessional › General knowledge work16991634−3.5
GPQA DiamondScience › Multidisciplinary96.392.7−3.5
OTIS Mock AIME 2024-2025Math100.098.1−1.9
HLAB-Vals-1.1Professional › Legal19.618.3−1.2
LMArena Agent Arena Tool HallucinationTool & computer use › Tool calling1.01.2+0.1
Data: Artificial Analysis, Vals AI, Surge AI and 9 othersatlas.village.dev
Win counts use each benchmark’s declared score direction. Represented suite slices and the derived Capability score are excluded from headline totals. Missing or incompatible setups are excluded, never counted as losses.
Loading Atlas data…