Atlas

Compare Models

Grok 4.6 vs Qwen3.8 Max (0902)

Qwen3.8 Max (0902) leads by 0.3 capability points (≈95% interval −1.9 to +2.5): statistically tied · P(higher) ≈ 61%

Wins 26 of 43 shared benchmarks (17 lost)

Capability vs. relative cost
Grok 4.6Qwen3.8 Max (0902)Closed weightsOpen weights

Each line runs through one model’s priced reasoning settings, cheapest to most expensive; grey points are other models’ priced settings in the same range. Results whose reasoning effort the source didn’t report are left out wherever a model has named effort levels. Capability and relative cost are the published estimates (relative cost is cost per task as a multiple of GPT-5 (high)); hover a setting, or use the arrow keys, to set it beside the other model’s same setting.

Grok 4.6vsQwen3.8 Max (0902)capability vs. relative costData Nov 15, 2026Village LLM Atlasatlas.village.dev

Benchmark wins17 – 26 · 43 shared

Comparing matching setups only (81 benchmarks excluded) · suites counted once
← Grok 4.6Qwen3.8 Max (0902) →

Grok 4.6vsQwen3.8 Max (0902)wins by categoryData Nov 15, 2026Data: Artificial Analysis, AIMING Lab (UNC-Chapel Hill), BuseyBench and 8 othersVillage LLM Atlasatlas.village.dev

Margins

Each benchmark scaled 0–100% (its declared range, else the field’s range); Δ is Qwen3.8 Max (0902) minus Grok 4.6.

AA-PersonalityBench Organization indexChat & Writing14.371.4+57.7
AA-PersonalityBench Conviction indexChat & Writing92.435.3−57.7
AA-PersonalityBench Expressiveness indexChat & Writing63.823.8−40.4
AA-PersonalityBench Slop indexChat & Writing35.373.3+38.5
AA-PersonalityBench Imagination indexChat & Writing20.048.6+28.8
AA-PersonalityBench Transparency indexChat & Writing71.444.8−26.9
Terminal-Bench 4.0 (Artificial Analysis)Code21.238.9+17.7
AA-PersonalityBench Warmth indexChat & Writing31.548.6+17.3
Co-Design BenchCode51.165.3+14.2
BuseyBench SVGMultimodal4.65.8+11.8
CritPt v2 (agentic, avg@5)Science62.351.4−10.9
AutomationBench-AA 1.0.6Agents67.056.2−10.8
Veracode GenAI Code Security (Fall 2026)Cybersecurity58.068.0+10.0
AA-Omniscience IndexIndexes30.512.0−9.2
Medical Long Context Reasoning (MLCR-AA)Health & Medicine12.220.0+7.8
FrontierSWE V2Code25.317.8−7.5
AA-PersonalityBench Informativeness indexChat & Writing44.839.1−5.8
Terminal-Bench-Science 0.1Science6.211.9+5.7
GDP.pdf (Artificial Analysis, All-pass)Professional Work17.822.8+5.0
SciCode 1.0.1 (Artificial Analysis)Code56.552.1−4.4
TCSExamMath40.544.8+4.3
AA-AnalystAgent (pass^5)Professional Work41.345.0+3.8
FrontierCode 2.0 (Artificial Analysis)Code19.823.4+3.7
DAXBench v2Code86.283.0−3.2
GDPval-AA v2Professional Work17531689−3.1
𝜏³-BankingAgents50.747.8−2.9
Artificial Analysis Intelligence Index v5.0.1Indexes31.033.7+2.6
AA-Briefcase EloAgents15771622+2.6
RSI-Exam 0.1 (Full 88)Agents0.40.4+2.5
Humanity's Last Exam DiamondGeneral QA21.524.0+2.5
MMMU ProMultimodal80.582.8+2.3
GPQA Diamond (Artificial Analysis avg@5)General QA94.992.8−2.1
CritPt (avg@5)Science19.717.7−2.0
Epoch Capabilities Index (ECI)Indexes158156−1.7
Terminal-Bench 4.0Code25.327.0+1.7
Artificial Analysis Intelligence Index v4.3.2Indexes44.345.4+1.1
Artificial Analysis Intelligence Index v4.3Indexes44.445.4+1.0
Artificial Analysis Progress Index v1Indexes322324+0.7
EnterpriseOps-Gym-AAAgents48.347.6−0.7
GDPval-AA v2.1Professional Work16581674+0.7
AA-LCR v1.1Search81.080.3−0.7
Terminal-Bench 2.1Code88.488.8+0.4
Humanity's Last Exam (Text-Only)General QA42.943.1+0.2

Grok 4.6vsQwen3.8 Max (0902)normalised score marginsData Nov 15, 2026Data: Artificial Analysis, AIMING Lab (UNC-Chapel Hill), BuseyBench and 8 othersVillage LLM Atlasatlas.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…