Atlas

Compare Models

Gemini 4 Carbon vs Qwen3.8 Max (0902)

Gemini 4 Carbon leads by 7.8 capability points (≈95% interval +5.0 to +10.5): a clear lead over Qwen3.8 Max (0902) · P(higher) ≈ 100%

Wins 28 of 31 shared benchmarks (3 lost)

Capability vs. relative cost
Gemini 4 CarbonQwen3.8 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 CarbonvsQwen3.8 Max (0902)capability vs. relative costData Nov 17, 2026Village LLM Atlasatlas.village.dev

Benchmark wins28 – 3 · 31 shared

Comparing matching setups only (123 benchmarks excluded) · suites counted once
← Gemini 4 CarbonQwen3.8 Max (0902) →

Gemini 4 CarbonvsQwen3.8 Max (0902)wins by categoryData Nov 17, 2026Data: Artificial Analysis, Proximal, BuseyBench and 2 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 Carbon.

Terminal-Bench-Science 0.1Science62.411.9−50.5
FrontierSWE V2 Worst@5Code56.28.1−48.1
Humanity's Last Exam DiamondGeneral QA69.124.0−45.1
FrontierSWE V2 Best@5Code66.726.8−39.9
VIALS-AA (pass^3)Multimodal64.031.7−32.3
CritPt v2 (agentic, avg@5)Science82.351.4−30.9
Artificial Analysis Intelligence Index v5.0.1Indexes60.033.7−26.4
Artificial Analysis Engineering IndexIndexes65.339.3−26.0
Artificial Analysis Intelligence Index v5.0Indexes57.733.6−24.2
Terminal-Bench 4.0 (Artificial Analysis)Code62.638.9−23.7
AutomationBench-AA 1.0.6Agents77.856.2−21.6
AA-Omniscience v1.1 IndexIndexes53.510.5−21.5
Artificial Analysis Legal IndexIndexes62.041.0−20.9
BuseyBench SVGMultimodal7.95.8−20.7
Artificial Analysis Finance & Accounting IndexIndexes62.645.0−17.6
FrontierCode 2.0 (Artificial Analysis)Code40.623.4−17.1
Artificial Analysis Economics IndexIndexes59.642.5−17.1
MKA (Artificial Analysis)Professional Work54.337.4−16.9
VIALS-AA (average of 3)Multimodal67.350.9−16.4
Artificial Analysis Healthcare & Medical IndexIndexes56.840.8−15.9
Artificial Analysis Strategy & Ops IndexIndexes60.947.3−13.7
Epoch Capabilities Index (ECI)Indexes170156−11.0
DAXBench v2Code93.683.0−10.6
Artificial Analysis Progress Index v1Indexes351324−9.5
GDP.pdf (Artificial Analysis, All-pass)Professional Work29.622.8−6.8
MMMU ProMultimodal89.282.8−6.5
GDPval-AA v2.1Professional Work17331674−2.6
VIALS-AA (pass@3)Multimodal70.872.7+1.9
AA-LCR v1.1Search79.380.3+1.0
Medical Long Context Reasoning (MLCR-AA, October 2026 methodology)Health & Medicine19.420.0+0.6
AA-Briefcase v1.1 EloAgents16111604−0.4

Gemini 4 CarbonvsQwen3.8 Max (0902)normalised score marginsData Nov 17, 2026Data: Artificial Analysis, Proximal, BuseyBench and 2 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…