Atlas

Compare Models

Claude Opus 5 vs Qwen3.8-Max

Claude Opus 5 leads by 4.4 capability points (≈95% interval +1.9 to +6.9): a clear lead over Qwen3.8-Max · P(higher) ≈ 100%

Wins 42 of 48 shared benchmarks (6 lost)

Capability vs. relative cost
Claude Opus 5Qwen3.8-MaxClosed 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 5vsQwen3.8-Maxcapability vs. relative costData Nov 15, 2026Village LLM Atlasatlas.village.dev

Benchmark wins42 – 6 · 48 shared

Comparing matching setups only (270 benchmarks excluded) · suites counted once
← Claude Opus 5Qwen3.8-Max →

Claude Opus 5vsQwen3.8-Maxwins by categoryData Nov 15, 2026Data: Artificial Analysis, Alibaba, Conceptual Reasoning Benchmark Authors and 17 othersVillage LLM Atlasatlas.village.dev

Margins

Each benchmark scaled 0–100% (its declared range, else the field’s range); Δ is Qwen3.8-Max minus Claude Opus 5.

BabaIsBench Stage 0: The IntroGames100.058.0−42.0
Medical Long Context Reasoning (MLCR-AA)Health & Medicine59.419.4−40.0
GameBench 2Games84.851.4−33.4
Terminal-Bench 4.0 (Artificial Analysis)Code49.018.7−30.3
OSWorld 2.0 - reported partial reward, patch and step limit unspecifiedAgents75.446.7−28.7
Continual Learning Exam (Trial 1)Agents42.218.2−24.0
Conceptual Reasoning IndexIndexes74.759.4−21.0
OSWorld 2.0 - reported binary completion, patch and step limit unspecifiedAgents39.619.4−20.2
NL2Repo - release-reported repository generation scoreCode75.355.9−19.4
AA-Briefcase EloAgents17201388−19.3
LMCAGeneral QA64.546.2−18.2
DeepsecBench — Revalidated V1Code32.415.5−16.9
DeepsecBench — Revalidated V1 (233 issues)Code32.315.5−16.8
AA-Omniscience IndexIndexes37.13.4−16.8
DeepsecBenchCode32.616.5−16.0
RSI-Exam 0.1 (Full 88)Agents0.50.3−14.2
ACCoRDGeneral QA83.871.4−12.4
Humanity's Last Exam (Text-Only)General QA54.443.0−11.4
GDPval-AA v2Professional Work18611630−11.3
Artificial Analysis Intelligence Index v4.3.2Indexes50.840.2−10.6
Artificial Analysis Intelligence Index v4.3Indexes50.740.3−10.4
CritPt (avg@5)Science29.120.0−9.1
BuseyBench SVGMultimodal6.35.4−8.8
Finance Agent v2Finance & Accounting58.650.6−8.0
Terminal-Bench 2.1Code89.181.3−7.8
Humanity's Last Exam (with tools)General QA63.656.2−7.4
AutomationBench-AA 1.0.6Agents56.649.2−7.3
𝜏³-BankingAgents44.751.3+6.6
YC-BenchGames19115922095838+6.4
LVBenchMultimodal75.481.8+6.4
DTBenchGeneral QA97.992.0−5.9
Epoch Capabilities Index (ECI)Indexes164157−5.7
GDPval-AA v2.1Professional Work17291611−5.2
Artificial Analysis Progress Index v1Indexes334320−5.0
CharXiv ReasoningMultimodal83.788.4+4.7
AA-LCR v1.1Search82.078.3−3.7
SciCode 1.0.1 (Artificial Analysis)Code56.453.2−3.1
ZeroBench (Main questions, pass@5, with tools)Multimodal52.049.0−3.0
BabyVision with toolsMultimodal94.191.3−2.8
OSWorld-VerifiedAgents83.486.1+2.7
FrontierChallenge (Pass Rate)Science17.515.5−2.1
MMMU ProMultimodal84.082.3−1.7
Agents' Last Exam - release-reported pass rate, subset unspecifiedAgents28.627.0−1.6
SCI-AGENT-EXAMScience64.262.7−1.5
GDP.pdf (Artificial Analysis, All-pass)Professional Work21.620.2−1.4
MRCR v2 (8-needle, Context Arena) - 128K AUCGeneral QA97.596.1−1.4
GPQA Diamond (Artificial Analysis avg@5)General QA93.792.7−1.0
MRCR v2 (8-needle, Context Arena) - 128K pointwiseGeneral QA91.392.3+1.0

Claude Opus 5vsQwen3.8-Maxnormalised score marginsData Nov 15, 2026Data: Artificial Analysis, Alibaba, Conceptual Reasoning Benchmark Authors and 17 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…