Atlas

Compare Models

Gemini 4 Carbon vs Qwen4-Max

Gemini 4 Carbon leads by 4.0 capability points (≈95% interval +1.4 to +6.7): a clear lead over Qwen4-Max · P(higher) ≈ 100%

Wins 40 of 50 shared benchmarks (9 lost, 1 tied)

Capability vs. relative cost
Gemini 4 CarbonQwen4-MaxClosed weightsOpen weights

Benchmark wins40 – 9 · 50 shared, 1 tied

Comparing matching setups only (4 benchmarks excluded) · suites counted once
← Gemini 4 CarbonQwen4-Max →
Data: Artificial Analysis, Surge AI, Vals AI and 4 othersatlas.village.dev

Margins

Each benchmark scaled 0–100% (its declared range, else the field’s range); Δ is Qwen4-Max minus Gemini 4 Carbon.

Terminal-Bench-Science 0.1Science › Multidisciplinary62.423.3−39.0
Humanity's Last Exam DiamondKnowledge & reasoning › General knowledge69.144.9−24.2
Code Migration v1.1 (Vals Index v2.2 subset)Software › Software engineering80.157.3−22.8
LMArena Agent Arena Net Improvement (Steering Burden)Tool & computer use › General agents16.17.3−17.2
Artificial Analysis Intelligence Index v5.0.1Composite indexes60.042.9−17.1
CritPt v2 (agentic, avg@5)Science › Physical sciences82.366.0−16.3
Artificial Analysis Engineering IndexScience › Engineering65.350.1−15.3
GDP.xlsxProfessional › General knowledge work38.923.7−15.2
sudo L7 (October 2026 rescoring)Software › Software engineering30.615.6−15.0
Terminal-Bench 4.0Software › Software engineering63.148.5−14.6
Legal Research BenchProfessional › Legal55.841.3−14.4
Terminal-Bench 4.0 (Artificial Analysis)Software › Software engineering62.648.5−14.1
FrontierCode 2.0 (Artificial Analysis)Software › Software engineering40.626.5−14.1
LMArena Agent Arena Praise vs ComplaintTool & computer use › General agents39.413.0−13.2
AA-Omniscience v1.1 IndexKnowledge & reasoning › General knowledge53.527.7−12.9
GDP.pdf (November 2026 rescoring)Professional › General knowledge work56.443.6−12.8
Artificial Analysis Legal IndexProfessional › Legal62.049.6−12.4
ComplexConstraints (November 2026 rescoring)Language & writing › Instruction following70.959.1−11.8
EnterpriseBench: CoreCraftTool & computer use › Customer service & operations80.568.7−11.8
KradleBenchGames22.011.0−11.0
DAYJOB: Healthcare (October 2026 rescoring)Health & medicine35.224.4−10.8
HANDBOOK.md Agents v1.1Tool & computer use › Customer service & operations49.839.4−10.4
LiveBench 2.0 (2026-11-06 release)Composite indexes83.373.1−10.3
Chartography (November 2026 rescoring)Perception & media › Documents & charts85.575.3−10.2
Artificial Analysis Healthcare & Medical IndexHealth & medicine56.848.2−8.6
Artificial Analysis Finance & Accounting IndexProfessional › Finance, accounting & tax62.654.1−8.5
Artificial Analysis Economics IndexComposite indexes59.651.3−8.3
Tax Agent BenchProfessional › Finance, accounting & tax76.470.0−6.4
Excel Modeling Benchmark v1.1Professional › Finance, accounting & tax76.570.8−5.6
AutomationBench-AA 1.0.6Tool & computer use › Customer service & operations77.872.6−5.3
Vals Legal IndexProfessional › Legal75.970.7−5.2
Vals Finance IndexProfessional › Finance, accounting & tax67.062.0−5.1
LMArena Agent Arena Steering BurdenTool & computer use › General agents17.98.0−4.9
LMArena Agent Arena Bash ReliabilityTool & computer use › General agents-5.73.4+4.5
LMArena Agent Arena Confirmed SuccessTool & computer use › General agents17.28.8−4.2
AA-Briefcase v1.1 EloProfessional › General knowledge work16111686+4.1
Artificial Analysis Strategy & Ops IndexProfessional › General knowledge work60.956.9−4.1
DAYJOB: TaxProfessional › Finance, accounting & tax4.88.8+4.0
HLAB-Vals-1.1Professional › Legal17.521.3+3.8
DAYJOB: Law (November 2026 release)Professional › Legal13.817.5+3.7
GDP.pdf (Artificial Analysis, All-pass)Professional › General knowledge work29.626.2−3.4
GDPval-AA v2.1Professional › General knowledge work17331809+3.4
DAYJOB: AccountingProfessional › Finance, accounting & tax25.222.4−2.8
LMArena Code Arena WebDevSoftware › Frontend & design17691750−2.5
Vibe Code Bench v1.1Software › Frontend & design89.186.7−2.4
DAYJOB: Finance (October 2026 rescoring)Professional › Finance, accounting & tax23.825.3+1.5
Finance Agent v2.1Professional › Finance, accounting & tax67.266.7−0.5
LLM Debate BenchmarkLanguage & writing › Language & wordplay16761679+0.4
LMArena Agent Arena Tool HallucinationTool & computer use › Tool calling0.81.2+0.2
Medical Long Context Reasoning (MLCR-AA, October 2026 methodology)Health & medicine19.419.40.0
Data: Artificial Analysis, Surge AI, Vals AI and 4 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…