Atlas

Compare Models

Gemini 4 Argon vs Qwen3.8 Max (0902)

Gemini 4 Argon leads by 5.6 capability points (≈95% interval +3.1 to +8.1): a clear lead over Qwen3.8 Max (0902) · P(higher) ≈ 100%

Wins 35 of 45 shared benchmarks (9 lost, 1 tied)

Capability vs. relative cost
Gemini 4 ArgonQwen3.8 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.

Gemini 4 ArgonvsQwen3.8 Max (0902)capability vs. relative costData Nov 15, 2026Village LLM Atlasatlas.village.dev

Benchmark wins35 – 9 · 45 shared, 1 tied

Comparing matching setups only (177 benchmarks excluded) · suites counted once
← Gemini 4 ArgonQwen3.8 Max (0902) →

Gemini 4 ArgonvsQwen3.8 Max (0902)wins by categoryData Nov 15, 2026Data: Artificial Analysis, Proximal, AIMING Lab (UNC-Chapel Hill) 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 Gemini 4 Argon.

AA-PersonalityBench Expressiveness indexChat & Writing84.823.8−61.5
AA-PersonalityBench Informativeness indexChat & Writing82.939.1−44.2
FrontierSWE V2 Best@5Code64.326.8−37.5
Terminal-Bench-Science 0.1Science48.611.9−36.7
FrontierSWE V2 Worst@5Code44.58.1−36.4
Terminal-Bench 2.1Code57.788.8+31.1
AA-PersonalityBench Warmth indexChat & Writing79.148.6−30.8
Terminal-Bench 4.0Code57.427.0−30.4
AA-PersonalityBench Conviction indexChat & Writing63.835.3−28.8
Last Translation Benchmark (LTBv1-eval, blind)Chat & Writing40.912.5−28.4
AA-3D v1.0 EloCode1147954−28.0
VIALS-AA (pass^3)Multimodal59.631.7−28.0
AA-PersonalityBench Transparency indexChat & Writing21.944.8+23.1
Visa Agentic Commerce BenchmarkAgents82.260.0−22.2
CritPt v2 (agentic, avg@5)Science72.951.4−21.4
AutomationBench-AA 1.0.6Agents77.556.2−21.3
Code2Video-AA EloCode12591021−19.4
Humanity's Last Exam DiamondGeneral QA42.524.0−18.5
BuseyBench SVGMultimodal7.65.8−17.7
Artificial Analysis Intelligence Index v5.0.1Indexes50.433.7−16.8
Terminal-Bench 4.0 (Artificial Analysis)Code55.138.9−16.2
AA-PersonalityBench Organization indexChat & Writing56.271.4+15.4
VIALS-AA (average of 3)Multimodal66.350.9−15.3
AA-Omniscience IndexIndexes42.412.0−15.2
Humanity's Last Exam (Text-Only)General QA57.143.1−14.0
MKA (Artificial Analysis)Professional Work51.337.4−13.8
RSI-Exam 0.1 (Full 88)Agents0.50.4−10.2
SciCode 1.0.1 (Artificial Analysis)Code61.852.1−9.7
CritPt (avg@5)Science27.117.7−9.4
Epoch Capabilities Index (ECI)Indexes167156−8.6
Artificial Analysis Intelligence Index v4.3.2Indexes52.645.4−7.1
AA-Briefcase v1.1 EloAgents14851604+6.5
Artificial Analysis Progress Index v1Indexes342324−6.4
Medical Long Context Reasoning (MLCR-AA, October 2026 methodology)Health & Medicine13.920.0+6.1
AA-PersonalityBench Slop indexChat & Writing67.673.3+5.8
MMMU ProMultimodal88.382.8−5.5
DAXBench v2Code88.383.0−5.3
AA-AnalystAgent (pass^5)Professional Work48.845.0−3.7
GDP.pdf (Artificial Analysis, All-pass)Professional Work25.622.8−2.8
FrontierCode 2.0 (Artificial Analysis)Code25.723.4−2.3
AA-PersonalityBench Imagination indexChat & Writing46.748.6+1.9
GDPval-AA v2.1Professional Work16341674+1.8
AA-LCR v1.1Search79.780.3+0.7
FrontierCode 1.1 Main (Artificial Analysis)Code43.543.2−0.3
VIALS-AA (pass@3)Multimodal72.772.70.0

Gemini 4 ArgonvsQwen3.8 Max (0902)normalised score marginsData Nov 15, 2026Data: Artificial Analysis, Proximal, AIMING Lab (UNC-Chapel Hill) 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…