Atlas

Compare Models

Claude Opus 5.5 vs Qwen3.8 Max (0902)

Claude Opus 5.5 leads by 7.4 capability points (≈95% interval +4.8 to +10.0): a clear lead over Qwen3.8 Max (0902) · P(higher) ≈ 100%

Wins 36 of 41 shared benchmarks (5 lost)

Capability vs. relative cost
Claude Opus 5.5Qwen3.8 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.

Claude Opus 5.5vsQwen3.8 Max (0902)capability vs. relative costData Nov 15, 2026Village LLM Atlasatlas.village.dev

Benchmark wins36 – 5 · 41 shared

Comparing matching setups only (113 benchmarks excluded) · suites counted once
← Claude Opus 5.5Qwen3.8 Max (0902) →

Claude Opus 5.5vsQwen3.8 Max (0902)wins by categoryData Nov 15, 2026Data: Artificial Analysis, Proximal, AIMING Lab (UNC-Chapel Hill) and 10 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 Claude Opus 5.5.

AA-PersonalityBench Expressiveness indexChat & Writing90.523.8−67.3
AA-PersonalityBench Slop indexChat & Writing20.073.3+53.8
AA-PersonalityBench Informativeness indexChat & Writing90.539.1−51.9
Terminal-Bench-Science 0.1Science61.911.9−50.0
Medical Long Context Reasoning (MLCR-AA)Health & Medicine66.720.0−46.7
AA-PersonalityBench Imagination indexChat & Writing94.348.6−46.2
AA-PersonalityBench Transparency indexChat & Writing90.544.8−46.2
FrontierSWE V2 Best@5Code71.526.8−44.7
FrontierSWE V2 Worst@5Code52.08.1−43.9
Code2Video-AA EloCode15361021−41.8
Terminal-Bench 4.0Code64.827.0−37.9
AA-PersonalityBench Warmth indexChat & Writing84.848.6−36.5
AA-3D v1.0 EloCode1202954−36.1
AA-PersonalityBench Conviction indexChat & Writing69.535.3−34.6
Humanity's Last Exam DiamondGeneral QA54.624.0−30.6
CritPt v2 (agentic, avg@5)Science81.751.4−30.3
TCSExam LiteMath72.345.4−26.9
FrontierCode 2.0 (Artificial Analysis)Code46.423.4−23.0
SPU (Science Prediction Universe)Science-0.30.1+22.0
Terminal-Bench 4.0 (Artificial Analysis)Code59.638.9−20.7
AA-Omniscience IndexIndexes46.412.0−17.2
SciCode 1.0.1 (Artificial Analysis)Code66.952.1−14.8
Humanity's Last Exam (Text-Only)General QA57.543.1−14.4
RSI-Exam 0.1 (Full 88)Agents0.50.4−14.3
CritPt (avg@5)Science31.717.7−14.0
FrontierCode 1.1 Main (Artificial Analysis)Code57.143.2−13.9
DAXBench v2Code95.783.0−12.8
AA-Briefcase v1.1 EloAgents18221604−12.0
AA-AnalystAgent (pass^5)Professional Work56.345.0−11.2
BuseyBench SVGMultimodal6.85.8−10.6
Epoch Capabilities Index (ECI)Indexes169156−10.4
GDPval-AA v2.1Professional Work18661674−8.5
AA-PersonalityBench Organization indexChat & Writing79.171.4−7.7
GDP.pdf (Artificial Analysis, All-pass)Professional Work28.822.8−6.0
Terminal-Bench 2.1Code83.588.8+5.3
Veracode GenAI Code Security (Fall 2026)Cybersecurity73.068.0−5.0
AA-LCR v1.1Search84.780.3−4.3
MMMU ProMultimodal86.682.8−3.8
Harvey LAB-AA v1.0 (criterion pass rate)Legal91.293.6+2.4
Visa Agentic Commerce BenchmarkAgents62.260.0−2.2
ITBench-AAAgents38.240.3+2.0

Claude Opus 5.5vsQwen3.8 Max (0902)normalised score marginsData Nov 15, 2026Data: Artificial Analysis, Proximal, AIMING Lab (UNC-Chapel Hill) and 10 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…