Atlas

Compare Models

Gemini 3.7 Flash vs Qwen3.8 Max (0902)

Qwen3.8 Max (0902) leads by 0.9 capability points (≈95% interval −1.3 to +3.1): statistically tied · P(higher) ≈ 80%

Wins 15 of 28 shared benchmarks (13 lost)

Capability vs. relative cost
Gemini 3.7 FlashQwen3.8 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 3.7 FlashvsQwen3.8 Max (0902)capability vs. relative costData Nov 15, 2026Village LLM Atlasatlas.village.dev

Benchmark wins13 – 15 · 28 shared

Comparing matching setups only (78 benchmarks excluded) · suites counted once
← Gemini 3.7 FlashQwen3.8 Max (0902) →

Gemini 3.7 FlashvsQwen3.8 Max (0902)wins by categoryData Nov 15, 2026Data: Artificial Analysis, AIMING Lab (UNC-Chapel Hill), BuseyBench and 6 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 3.7 Flash.

AA-Briefcase EloAgents11161622+29.4
Terminal-Bench 4.0 (Artificial Analysis)Code13.638.9+25.3
Terminal-Bench 4.0Code6.127.0+20.9
AA-AnalystAgent (pass^5)Professional Work60.045.0−15.0
GDPval-AA v2Professional Work14331689+12.6
𝜏³-BankingAgents35.547.8+12.4
GDPval-AA v2.1Professional Work14041674+11.9
BuseyBench SVGMultimodal6.65.8−8.5
RSI-Exam 0.1 (Full 88)Agents0.30.4+8.3
Veracode GenAI Code Security (Fall 2026)Cybersecurity60.068.0+8.0
SciCode 1.0.1 (Artificial Analysis)Code59.852.1−7.8
AA-Omniscience IndexIndexes26.512.0−7.2
Humanity's Last Exam DiamondGeneral QA31.224.0−7.2
Artificial Analysis Intelligence Index v4.3.2Indexes39.145.4+6.4
Artificial Analysis Intelligence Index v4.3Indexes39.445.4+6.0
AutomationBench-AA 1.0.6Agents62.056.2−5.8
Medical Long Context Reasoning (MLCR-AA)Health & Medicine15.020.0+5.0
Humanity's Last Exam (Text-Only)General QA39.043.1+4.1
CritPt (avg@5)Science14.317.7+3.4
Terminal-Bench 2.1Code85.888.8+3.0
EnterpriseOps-Gym-AAAgents50.447.6−2.8
AA-LCR v1.1Search83.080.3−2.7
FrontierSWE V2Code20.317.8−2.5
Artificial Analysis Progress Index v1Indexes317324+2.2
Epoch Capabilities Index (ECI)Indexes158156−2.1
MMMU ProMultimodal84.982.8−2.1
GPQA Diamond (Artificial Analysis avg@5)General QA94.592.8−1.7
GDP.pdf (Artificial Analysis, All-pass)Professional Work23.622.8−0.8

Gemini 3.7 FlashvsQwen3.8 Max (0902)normalised score marginsData Nov 15, 2026Data: Artificial Analysis, AIMING Lab (UNC-Chapel Hill), BuseyBench and 6 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…