Enter opens the highlighted result. Esc closes.

Claude Opus 5 vs Qwen4 Max

Claude Opus 5 leads Qwen4 Max on coding, 81.7 to 74.0 on the Vibe Coding Index, and Qwen4 Max costs 5.8x less per output token ($4.275 vs $25.00 per million). Updated November 22, 2026; 17 benchmarks have a result for both.

Claude Opus 5Anthropic
Coding
81.7
Overall
80.1
Output $/M
$25.00
11
Qwen4 MaxQwen
Coding
74.0
Overall
66.6
Output $/M
$4.275
6
vs boards won of 17 shared

Claude Opus 5 or Qwen4 Max: which to pick

Pick Claude Opus 5 for coding: it scores 81.7 on the Vibe Coding Index against 74.0 for Qwen4 Max.

Pick Qwen4 Max for output tokens 5.8x cheaper ($4.275 vs $25.00 per million) and work measured by LMArena WebDev or LMArena WebDev: React, where it leads.

Shared results
17
Claude Opus 5 only
125
Qwen4 Max only
9
Like-for-like categories
6 of 9

Only benchmarks that tested both models count toward a pick; a result only one model has is shown but not compared.

Which one for your work

Each card is decided by one number both models carry. Where only one has it, the card says so and makes no call.

At a glance

Coding index
Claude Opus 5 81.7 from 5 benchmarks, Qwen4 Max 74.0 from 4 benchmarks.
Widest category
Agentic work, 10.3 points apart on average over 2 shared benchmarks.
Widest single gap
Legal Research Bench (Vals AI): Claude Opus 5 55.3% against Qwen4 Max 41.4%, 13.9 points apart on a 0 to 100 scale.
Output price
Claude Opus 5 $25.00 against Qwen4 Max $4.275 per million output tokens, 5.8x apart.
Context window
Both take 1M tokens.

Benchmark by benchmark

First the score: boards each model leads, per category, where both have a result. Then every board, a category to a fold; bars are each result on a 0 to 100 scale, so compare the two in a row, not one row with another.

Boards won, where both have a result
CategoryClaude Opus 5BarsQwen4 MaxTies
Coding64
Intelligence10
Agentic work20
Reasoning11
Knowledge10
General01
All shared boards116
CodingLike-for-likeSix of these boards feed the Vibe Coding Index64
Coding benchmarks for Claude Opus 5 and Qwen4 Max
BenchmarkClaude Opus 5Qwen4 MaxLeader
LiveBench codingNot measured69.8%Only Qwen4 Max measured
LMArena WebDev16921751Qwen4 Max
LiveCodeBench (Vals AI)89.0%Not measuredOnly Claude Opus 5 measured
SWE-bench Verified (Vals AI)97.0%Not measuredOnly Claude Opus 5 measured
Terminal-Bench 4 (Vals AI)53.5%48.5%Claude Opus 5
Vibe Code Bench (Vals AI)88.4%86.7%Claude Opus 5
Code migration (Vals AI)57.5%Not measuredOnly Claude Opus 5 measured
ProgramBench (Vals AI)3.0%Not measuredOnly Claude Opus 5 measured
IOI (Vals AI)84.3%Not measuredOnly Claude Opus 5 measured
Terminal-Bench 2.1 (Vals AI)84.6%Not measuredOnly Claude Opus 5 measured
Vibe Code Bench 1-100 (Vals AI)28.5%Not measuredOnly Claude Opus 5 measured
SkillsBench (Vals AI)60.4%Not measuredOnly Claude Opus 5 measured
CyberBench (Vals AI)65.4%Not measuredOnly Claude Opus 5 measured
SciCode56.4%Not measuredOnly Claude Opus 5 measured
Terminal-Bench 4.0 (Artificial Analysis run)49.0%48.5%Claude Opus 5
FrontierCode 2.031.2%26.5%Claude Opus 5
FrontierCode53.4%Not measuredOnly Claude Opus 5 measured
FrontierSWE52.0%Not measuredOnly Claude Opus 5 measured
WeirdML91.8%Not measuredOnly Claude Opus 5 measured
MirrorCode87.8%Not measuredOnly Claude Opus 5 measured
ALE-Bench2164.61831.0Claude Opus 5
LMArena Text: coding1532Not measuredOnly Claude Opus 5 measured
LMArena Text: software and IT1526Not measuredOnly Claude Opus 5 measured
LMArena WebDev: React17131785Qwen4 Max
LMArena WebDev: HTML16811752Qwen4 Max
LMArena WebDev: from an image16641675Qwen4 Max
SciCode13.8%Not measuredOnly Claude Opus 5 measured
SWE-Bench Pro V299.4%Not measuredOnly Claude Opus 5 measured
SWE Atlas: Codebase QnA63.2%Not measuredOnly Claude Opus 5 measured
SWE Atlas: Test Writing62.2%Not measuredOnly Claude Opus 5 measured
GameBench 284.882.4Claude Opus 5
Terminal-Bench 3.042.7%Not measuredOnly Claude Opus 5 measured
Terminal-Bench 4.053.9%Not measuredOnly Claude Opus 5 measured
ProgramBench4.5%Not measuredOnly Claude Opus 5 measured
Neither model has a result on the other 18 boards here
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index10
Intelligence benchmarks for Claude Opus 5 and Qwen4 Max
BenchmarkClaude Opus 5Qwen4 MaxLeader
AA Intelligence Index44.242.9Claude Opus 5
Epoch Capabilities Index164.7Not measuredOnly Claude Opus 5 measured
Agentic workLike-for-likePublic boards, shown as published, kept out of the index20
Agentic work benchmarks for Claude Opus 5 and Qwen4 Max
BenchmarkClaude Opus 5Qwen4 MaxLeader
Design Arena: agents building web apps1236Not measuredOnly Claude Opus 5 measured
Design Arena: agents building full-stack apps1285Not measuredOnly Claude Opus 5 measured
Design Arena: agents building mobile apps1348Not measuredOnly Claude Opus 5 measured
Design Arena: agents building Android apps1243Not measuredOnly Claude Opus 5 measured
Design Arena: agents building Godot games1249Not measuredOnly Claude Opus 5 measured
Design Arena: agents building games1203Not measuredOnly Claude Opus 5 measured
Design Arena: PowerPoint slides1225Not measuredOnly Claude Opus 5 measured
Design Arena: spreadsheets1241Not measuredOnly Claude Opus 5 measured
LiveBench researchNot measured84.7%Only Qwen4 Max measured
LiveBench browserNot measured80.7%Only Qwen4 Max measured
LiveBench long agenticNot measured46.6%Only Qwen4 Max measured
SREBench (Vals AI)12.2%Not measuredOnly Claude Opus 5 measured
Finance Agent (Vals AI)58.6%Not measuredOnly Claude Opus 5 measured
Tax Agent Bench (Vals AI)47.1%40.4%Claude Opus 5
Legal Research Bench (Vals AI)55.3%41.4%Claude Opus 5
Harvey's Legal Agent Benchmark (Vals AI)Not measured21.3%Only Qwen4 Max measured
Excel Modeling Benchmark (Vals AI)73.6%Not measuredOnly Claude Opus 5 measured
BioMysteryBench (Vals AI)79.3%Not measuredOnly Claude Opus 5 measured
Terminal-Bench-Science (Vals AI)27.1%Not measuredOnly Claude Opus 5 measured
Vals Science Index47.9%Not measuredOnly Claude Opus 5 measured
Time Horizon Index (Vals AI)18.8%Not measuredOnly Claude Opus 5 measured
CUA-bench (Vals AI)9.0%Not measuredOnly Claude Opus 5 measured
APEX-Agents65.8%Not measuredOnly Claude Opus 5 measured
OSWorld 2.031.4%Not measuredOnly Claude Opus 5 measured
BALROG63.4%Not measuredOnly Claude Opus 5 measured
PostTrainBench35.7%Not measuredOnly Claude Opus 5 measured
Vending-Bench 2$11,182Not measuredOnly Claude Opus 5 measured
LMArena Search1238Not measuredOnly Claude Opus 5 measured
MCP Atlas85.8%Not measuredOnly Claude Opus 5 measured
τ³-Banking48.7%Not measuredOnly Claude Opus 5 measured
Neither model has a result on the other 14 boards here
ReasoningLike-for-likePublic boards, shown as published, kept out of the index11
Reasoning benchmarks for Claude Opus 5 and Qwen4 Max
BenchmarkClaude Opus 5Qwen4 MaxLeader
GPQA Diamond93.4%Not measuredOnly Claude Opus 5 measured
LiveBench reasoningNot measured92.2%Only Qwen4 Max measured
MysteryMechanism (Vals AI)37.4%Not measuredOnly Claude Opus 5 measured
MMMU Pro (Vals AI)89.9%Not measuredOnly Claude Opus 5 measured
GPQA Diamond (Artificial Analysis run)93.7%Not measuredOnly Claude Opus 5 measured
Humanity's Last Exam (Artificial Analysis run)54.9%Not measuredOnly Claude Opus 5 measured
HLE-Diamond (Artificial Analysis run)44.8%44.9%Qwen4 Max
CritPt v2, hard subset (Artificial Analysis run)50.0%43.0%Claude Opus 5
GPQA Diamond (Epoch AI run)93.9%Not measuredOnly Claude Opus 5 measured
ARC-AGI97.5%Not measuredOnly Claude Opus 5 measured
ARC-AGI-290.4%Not measuredOnly Claude Opus 5 measured
SimpleBench80.6%Not measuredOnly Claude Opus 5 measured
Chess puzzles43.0%Not measuredOnly Claude Opus 5 measured
Mystery game puzzles59.0%Not measuredOnly Claude Opus 5 measured
EBR-bench45.7%Not measuredOnly Claude Opus 5 measured
LMArena Text: hard prompts1525Not measuredOnly Claude Opus 5 measured
LMArena Text: expert1553Not measuredOnly Claude Opus 5 measured
LMArena Text: hard prompts in English1523Not measuredOnly Claude Opus 5 measured
LMArena Vision: diagrams1345Not measuredOnly Claude Opus 5 measured
LMArena Vision: homework1336Not measuredOnly Claude Opus 5 measured
Humanity's Last Exam (Scale run)38.6%Not measuredOnly Claude Opus 5 measured
GoBench2254Not measuredOnly Claude Opus 5 measured
Neither model has a result on the other 4 boards here
MathDirectional onlyPublic boards, shown as published, kept out of the index7 boards, none shared
Math benchmarks for Claude Opus 5 and Qwen4 Max
BenchmarkClaude Opus 5Qwen4 MaxLeader
OTIS Mock AIME98.9%Not measuredOnly Claude Opus 5 measured
FrontierMath, tiers 1 to 385.6%Not measuredOnly Claude Opus 5 measured
FrontierMath, tier 473.2%Not measuredOnly Claude Opus 5 measured
ProofBench99.0%Not measuredOnly Claude Opus 5 measured
LMArena Text: math1532Not measuredOnly Claude Opus 5 measured
LMArena Text: mathematical work1543Not measuredOnly Claude Opus 5 measured
MathArena64.1%Not measuredOnly Claude Opus 5 measured
Neither model has a result on the other 7 boards here
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index10
Knowledge benchmarks for Claude Opus 5 and Qwen4 Max
BenchmarkClaude Opus 5Qwen4 MaxLeader
MMLU Pro91.6%Not measuredOnly Claude Opus 5 measured
LegalBench (Vals AI)87.0%Not measuredOnly Claude Opus 5 measured
CorpFin (Vals AI)73.2%Not measuredOnly Claude Opus 5 measured
TaxEval (Vals AI)75.1%Not measuredOnly Claude Opus 5 measured
MortgageTax (Vals AI)72.1%Not measuredOnly Claude Opus 5 measured
MedCode (Vals AI)63.6%Not measuredOnly Claude Opus 5 measured
AA-Omniscience37.627.7Claude Opus 5
SimpleQA Verified59.9%Not measuredOnly Claude Opus 5 measured
LMArena Text: science1533Not measuredOnly Claude Opus 5 measured
LMArena Text: medicine and healthcare1526Not measuredOnly Claude Opus 5 measured
LMArena Text: legal and government1527Not measuredOnly Claude Opus 5 measured
LMArena Text: business and finance1499Not measuredOnly Claude Opus 5 measured
Neither model has a result on the other 9 boards here
Web designDirectional onlyPublic boards, shown as published, kept out of the index9 boards, none shared
Web design benchmarks for Claude Opus 5 and Qwen4 Max
BenchmarkClaude Opus 5Qwen4 MaxLeader
Design Arena: websites1299Not measuredOnly Claude Opus 5 measured
Design Arena: UI components1335Not measuredOnly Claude Opus 5 measured
Design Arena: game dev1316Not measuredOnly Claude Opus 5 measured
Design Arena: data viz1324Not measuredOnly Claude Opus 5 measured
Design Arena: 3D1309Not measuredOnly Claude Opus 5 measured
Design Arena: all code categories1310Not measuredOnly Claude Opus 5 measured
Design Arena: SVG1298Not measuredOnly Claude Opus 5 measured
Design Arena: ASCII art1338Not measuredOnly Claude Opus 5 measured
LiveBench designNot measured79.1%Only Qwen4 Max measured
GeneralLike-for-likePublic boards, shown as published, kept out of the index01
General benchmarks for Claude Opus 5 and Qwen4 Max
BenchmarkClaude Opus 5Qwen4 MaxLeader
Vals IndexNot measured63.2%Only Qwen4 Max measured
LiveBench data analysisNot measured83.4%Only Qwen4 Max measured
Public Benefits Bench (Vals AI)76.9%Not measuredOnly Claude Opus 5 measured
MedScribe (Vals AI)91.0%Not measuredOnly Claude Opus 5 measured
Vals Multimodal Index73.9%Not measuredOnly Claude Opus 5 measured
Long context reasoning (AA-LCR)82.0%Not measuredOnly Claude Opus 5 measured
GDPval-AA17291809Qwen4 Max
DTBench97.9%Not measuredOnly Claude Opus 5 measured
LMCA64.5%Not measuredOnly Claude Opus 5 measured
Furniture assembly60.8%Not measuredOnly Claude Opus 5 measured
LMArena Text1506Not measuredOnly Claude Opus 5 measured
LMArena Text: instruction following1517Not measuredOnly Claude Opus 5 measured
LMArena Text: creative writing1494Not measuredOnly Claude Opus 5 measured
LMArena Vision1317Not measuredOnly Claude Opus 5 measured
LMArena Text: English1507Not measuredOnly Claude Opus 5 measured
LMArena Text: non-English1500Not measuredOnly Claude Opus 5 measured
LMArena Text: Chinese1575Not measuredOnly Claude Opus 5 measured
LMArena Text: German1519Not measuredOnly Claude Opus 5 measured
LMArena Text: Japanese1522Not measuredOnly Claude Opus 5 measured
LMArena Text: Korean1503Not measuredOnly Claude Opus 5 measured
LMArena Text: French1521Not measuredOnly Claude Opus 5 measured
LMArena Text: Spanish1522Not measuredOnly Claude Opus 5 measured
LMArena Text: Russian1507Not measuredOnly Claude Opus 5 measured
LMArena Text: Polish1523Not measuredOnly Claude Opus 5 measured
LMArena Text: multi-turn1499Not measuredOnly Claude Opus 5 measured
LMArena Text: longer queries1516Not measuredOnly Claude Opus 5 measured
LMArena Text: writing and language1499Not measuredOnly Claude Opus 5 measured
LMArena Text: entertainment and media1480Not measuredOnly Claude Opus 5 measured
LMArena Vision: English1310Not measuredOnly Claude Opus 5 measured
LMArena Vision: Chinese1401Not measuredOnly Claude Opus 5 measured
LMArena Vision: reading text1325Not measuredOnly Claude Opus 5 measured
LMArena Vision: humor1293Not measuredOnly Claude Opus 5 measured
LMArena Vision: creative writing1308Not measuredOnly Claude Opus 5 measured
LMArena Document1509Not measuredOnly Claude Opus 5 measured
EQ-Bench 41365Not measuredOnly Claude Opus 5 measured
Neither model has a result on the other 14 boards here
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
Image generation benchmarks for Claude Opus 5 and Qwen4 Max
BenchmarkClaude Opus 5Qwen4 MaxLeader
Neither model has a result on 5 boards here

What each workload costs

Three fixed token mixes at each model's listed price, so per-token rates become a bill.

Chat turn

1K input + 500 output tokens

Claude Opus 5$0.018Fits in one request
Qwen4 Max$0.0036Fits in one request

Qwen4 Max costs less

Repository review

50K input + 3K output tokens

Claude Opus 5$0.325Fits in one request
Qwen4 Max$0.084Fits in one request

Qwen4 Max costs less

Agent loop

200K input + 10K output tokens

Claude Opus 5$1.25Fits in one request
Qwen4 Max$0.328Fits in one request

Qwen4 Max costs less

Listed OpenRouter rates, without cache or batch discounts. A workload that does not fit a model's context window would need splitting.

Specification differences

The leading value in a row is marked; a row with a missing value names no leader. Our two rows are marked in red.

Specifications of Claude Opus 5 and Qwen4 Max
SpecificationAnthropic Claude Opus 5Qwen4 Max
MakerAnthropicQwen
WeightsNo public weights listedNo public weights listed
ReleasedJuly 24, 2026November 16, 2026
Context window1M1M
Input $/Mtok$5.00$1.425
Output $/Mtok$25.00$4.275
Coding index81.774.0
Independent measurements behind it54
Cost per solved task$1.33Not computable
SourceOpenRouter listingOpenRouter listing

Questions

Is Claude Opus 5 better than Qwen4 Max for coding?

Pick Claude Opus 5 for coding: it scores 81.7 on the Vibe Coding Index against 74.0 for Qwen4 Max. Pick Qwen4 Max for output tokens 5.8x cheaper ($4.275 vs $25.00 per million) and work measured by LMArena WebDev or LMArena WebDev: React, where it leads.

Which is cheaper, Claude Opus 5 or Qwen4 Max?

Claude Opus 5 costs $5.00 per million input tokens and $25.00 per million output tokens; Qwen4 Max costs $1.425 and $4.275. Cost per solved task is known for Claude Opus 5 only.

Which has the bigger context window, Claude Opus 5 or Qwen4 Max?

Neither: both take 1M tokens.

How many benchmarks is the Claude Opus 5 vs Qwen4 Max comparison based on?

17 benchmarks have a result for both models as of November 22, 2026. Claude Opus 5's coding index rests on 5 independent benchmarks and Qwen4 Max's on 4. Where only one model has a result, the page says so instead of naming a winner.

Compare next

Embed this comparison on your site

Free under CC BY 4.0, while the source link stays visible. Preview the frame · Terms and other embeds

<iframe src="https://llmbenchmark.ai/embed/compare/claude-opus-5-vs-qwen4-max" title="Claude Opus 5 vs Qwen4 Max, LLMBenchmark.ai" width="100%" height="360" style="border:0;max-width:100%" loading="lazy"></iframe>
<p style="font:13px/1.4 sans-serif;margin:6px 0 0">Source: <a href="https://llmbenchmark.ai/compare/claude-opus-5-vs-qwen4-max">LLMBenchmark.ai, Claude Opus 5 vs Qwen4 Max</a></p>