BenchLeader
Menu
Best LLMsChoose a modelLeaderboardsModelsBenchmarksProvidersReleases9

Claude Opus 4.7 vs Qwen3.6 Max

Full profiles: Claude Opus 4.7 (Anthropic) and Qwen3.6 Max (Alibaba): every benchmark, price and rank.

Updated · Data from 9 sources

As of 26 Oct 2026, Claude Opus 4.7 and Qwen3.6 Max (max) are neck and neck on quality (63.2 vs 62.3 on the BenchLeader Index, inside the ±4.0 margin of error), and Qwen3.6 Max (max) costs 3.4× less, which makes it the better buy for most workloads. Pick Claude Opus 4.7 for reasoning (65 vs 58); pick Qwen3.6 Max for agents & tools (72 vs 65).

Claude Opus 4.7 benchmarks Anthropic Qwen3.6 Max benchmarks max AlibabaOpen in the comparison tool

Which to pick for…

The pick's numbers come first. Category scores on the BenchLeader scale (50 is the average model, 15 points one standard deviation). High confidence means a lead of half a standard deviation or more; under one point is called either.

Use casePickWhyConfidence
CodingClaude Opus 4.761 vs 57 category score; Arena WebDev 1555 vs 1482Medium
Agents & tool useQwen3.6 Max72 vs 65 category score; Terminal-Bench 2.0 (Vals) 51.7% vs 68.5%Medium
MathsClaude Opus 4.767 vs 65 category score; Arena Maths 1491 vs 1476Medium
ReasoningClaude Opus 4.765 vs 58 category score; Arena Hard Prompts 1519 vs 1483Medium
KnowledgeEither62 vs 62 category score; CorpFin 66.1% vs 66.5%Either
Chat quality (human preference)Claude Opus 4.768 vs 65 on Arena votesMedium
Lowest costQwen3.6 Max$2.96 vs $10.00 per 1M blended; agentic coding session $0.111 vs $0.400–
High volume on a budgetNeither flagshipA smaller model is usually the better fit: Cheapest good LLMs, Fastest LLMs–

Key differences

  • Claude Opus 4.7 and Qwen3.6 Max (max) are neck and neck on quality (63.2 vs 62.3, within the ±4.0 margin).
  • Claude Opus 4.7 is stronger in coding, human preference, maths, multimodal, reasoning.
  • Qwen3.6 Max (max) is stronger in agents & tools, composite, instruction following, long context.
  • Qwen3.6 Max (max) is 3.4× cheaper ($2.96 vs $10.00 per 1M blended).

At a glance

SpecClaude Opus 4.7Qwen3.6 Max
DeveloperAnthropicAlibaba
Released16 Apr 202620 Apr 2026
Context window (tokens)1,000,000245,800
Compared atClaude Opus 4.7Qwen3.6 Max (max)
Open weightsNoNo
Price in / out / cached input per 1M$5.00 / $25.00 / $0.500$1.32 / $7.90 / $0.132

Category by category

Category scores on the BenchLeader scale: 50 is the average evaluated model, 15 points is one standard deviation. A gap under one point is called level.

CategoryClaude Opus 4.7Qwen3.6 Max (max)Leader
Reasoning64.858.1Claude Opus 4.7
Coding61.456.6Claude Opus 4.7
Agents & tools65.472.1Qwen3.6 Max
Maths66.764.9Claude Opus 4.7
Knowledge62.362.3level
Instruction following–74.6not comparable
Human preference67.564.8Claude Opus 4.7
Multimodal66.8–not comparable
Long context–65.4not comparable
Composite–61.2not comparable

Where sources disagree

Across the 15 benchmarks both have, Claude Opus 4.7 leads 13, Qwen3.6 Max leads 2. Every overall index that rates both puts Claude Opus 4.7 ahead.

Overall indexClaude Opus 4.7Qwen3.6 Max (max)Leader
BenchLeader Index63.262.3neck and neck
AA Intelligence Index v4.3.230.9†28.4Claude Opus 4.7
Arena Text14941461Claude Opus 4.7

Each publisher runs its own tasks, harnesses and effort settings, so their overall verdicts can differ; BenchLeader reconciles them on one scale. † measured at another effort setting of the same model (the publisher did not test the one compared here).

Price, speed and latency

List prices blended at three input tokens per output token. Per-task prices use the token counts on the methodology page; suite cost is what Artificial Analysis paid to run its whole Intelligence Index.

MeasureClaude Opus 4.7Qwen3.6 Max (max)
Blended price per 1M tokens$10.00$2.96
Input / output per 1M$5.00 / $25.00$1.32 / $7.90
Cached input per 1M$0.500$0.132
Price per task: chat reply$0.0095$0.0029
Price per task: summarize a 30-page report$0.075$0.021
Price per task: code edit$0.068$0.020
Price per task: agentic coding session$0.400$0.111
Price per task: structured extraction$0.015$0.0042
Output speed60 tok/s–
Time to first answer1.2 s–

Each model is compared at its best-ranked reasoning-effort setting. Claude Opus 4.7 scores 63.2 and takes 1 s to its first answer token; Claude Opus 4.7 (max) scores 62.4 in 25 s, Claude Opus 4.7 (high) scores 61.0 in 1 s. Qwen3.6 Max (max) scores 62.3. Higher effort means more thinking before the answer: better scores, longer waits and more tokens billed. How reasoning effort changes scores and cost.

Every benchmark

105 benchmarks, grouped by category; each name links to the board and its leaders. † measured at another effort setting of the same model, shown for reference and not counted as a win.

MetricPublisherClaude Opus 4.7Qwen3.6 Max (max)
BenchLeader IndexBenchLeader63.262.3
Agents & tools scoreBenchLeader65.472.1
Coding scoreBenchLeader61.456.6
Human preference scoreBenchLeader67.564.8
Knowledge scoreBenchLeader62.362.3
Maths scoreBenchLeader66.764.9
Multimodal scoreBenchLeader66.8–
Reasoning scoreBenchLeader64.858.1
Composite scoreBenchLeader–61.2
Instruction following scoreBenchLeader–74.6
Long context scoreBenchLeader–65.4
Blended price $/MBenchLeader$10.00$2.96
Output speedBenchLeader60 tok/s–
Time to first answerBenchLeader1.2 s–
Context windowBenchLeader1,000,000 tokens245,800 tokens
Reasoning
GPQA DiamondEpoch AI Benchmarking Hub86.4%†87.4%
Humanity's Last ExamScale AI / CAIS36.2%–
SimpleBenchSimpleBench61.7%63.0%
Arena Hard PromptsArena (LMArena)15191483
LiveBench 1.0 ReasoningLiveBench87.2%†–
GPQA Diamond (AA)Artificial Analysis91.4%†88.8%
Humanity's Last Exam (AA)Artificial Analysis42.3%†30.8%
GPQA Diamond (Vals)Vals AI90.2%–
Kagi LLM BenchmarkKagi LLM Benchmark73.3%–
ARC-AGI-3ARC Prize0.2%†–
CritPtArtificial Analysis12.0%†3.7%
Chess PuzzlesEpoch AI Benchmarking Hub41.0%†–
Mystery Game PuzzlesEpoch AI Benchmarking Hub13.0%19.0%†
LMCAEpoch AI Benchmarking Hub52.2%†42.5%
DTBenchEpoch AI Benchmarking Hub94.7%†87.2%
ForecastBenchForecastBench63.1%–
Coding
SWE-bench Verified (Epoch)Epoch AI Benchmarking Hub83.5%†76.7%
SciCodeSciCode54.5%†–
WeirdMLWeirdML76.4%–
FrontierCodeCognition38.5%–
GSO-BenchGSO-Bench44.1%–
Arena CodingArena (LMArena)15461512
Arena WebDevArena (LMArena)15551482
LiveBench 1.0 CodingLiveBench82.1%†–
LiveCodeBenchVals AI85.1%–
Show all 105 benchmarks
MetricPublisherClaude Opus 4.7Qwen3.6 Max (max)
Coding
SWE-bench (Vals)Vals AI82.0%72.8%
Code MigrationVals AI43.9%–
ProgramBenchVals AI0.0%–
Vibe Code Bench v1.1Vals AI71.0%–
SWE Atlas: Codebase QnAScale AI SEAL40.3%–
SWE Atlas: RefactoringScale AI SEAL48.6%–
SWE Atlas: Test WritingScale AI SEAL38.5%–
MirrorCodeEpoch AI Benchmarking Hub31.1%†–
ALE-BenchEpoch AI Benchmarking Hub1323.0–
Terminal-Bench 2.1 (AA)Artificial Analysis83.2%†–
Agents & tools
Terminal-BenchTerminal-Bench80.2%–
OSWorld-Verified 2.0OSWorld18.2%†–
APEX-AgentsMercor49.2%†–
LiveBench Agentic CodingLiveBench50.7%†–
Terminal-Bench HardArtificial Analysis51.5%†43.9%
τ²-Bench Telecom (AA)Artificial Analysis88.6%†95.9%
Terminal-Bench 2.1 (Vals)Vals AI68.5%–
MCP AtlasScale AI SEAL79.1%†–
HiL-BenchScale AI SEAL41.7%–
τ³-bench Bankingτ²-bench40.2%†–
τ³-Banking (AA)Artificial Analysis34.6%†–
ITBench SRE (AA)Artificial Analysis46.7%†–
Analyst Agent (AA)Artificial Analysis43.8%†–
Finance Agent v2Vals AI51.5%–
Harvey's Legal Agent Benchmark v1Vals AI6.7%–
Legal Research BenchVals AI38.5%–
Tax Agent BenchVals AI23.7%–
Terminal-Bench 2.0 (Vals)Vals AI68.5%51.7%
Arena SearchArena (LMArena)1210–
PostTrainBenchEpoch AI Benchmarking Hub28.6%†–
ExploitBenchEpoch AI Benchmarking Hub26.5%–
GBAEvalEpoch AI Benchmarking Hub43.8%–
Vending-Bench 2Andon Labs10936.84254.2
Maths
FrontierMath Tiers 1–3Epoch AI Benchmarking Hub70.2%†–
FrontierMath Tier 4Epoch AI Benchmarking Hub31.7%†–
OTIS Mock AIMEEpoch AI Benchmarking Hub86.7%†91.1%
ProofBenchVals AI54.0%†–
LiveBench MathematicsLiveBench92.8%†–
AIME (Vals)Vals AI96.3%–
AIME 2026MathArena95.8%†–
HMMT February 2026MathArena93.9%†–
MathArena ApexMathArena40.6%†–
Arena MathsArena (LMArena)14911476
Knowledge
SimpleQA VerifiedEpoch AI Benchmarking Hub51.7%†52.0%
LiveBench 1.0 Data AnalysisLiveBench78.3%†–
AA-OmniscienceArtificial Analysis27.3†9.2
MMLU-ProVals AI89.9%–
LegalBenchVals AI85.3%–
CorpFinVals AI66.1%66.5%
TaxEvalVals AI75.3%–
CaseLaw v2Vals AI68.4%47.9%
Excel Modeling BenchmarkVals AI63.7%–
MedCodeVals AI54.9%–
MedScribeVals AI83.0%–
EBR-benchEpoch AI Benchmarking Hub19.1%†–
GDP.pdfEpoch AI Benchmarking Hub21.0%†–
AA-Omniscience: accuracyArtificial Analysis48.9%†37.9%
AA-Omniscience: non-hallucinationArtificial Analysis57.7%†53.8%
Instruction following
LiveBench LanguageLiveBench77.9%†–
LiveBench Instruction FollowingLiveBench66.7%†–
IFBenchArtificial Analysis58.6%†76.6%
Arena Instruction FollowingArena (LMArena)14931452
Human preference
Arena TextArena (LMArena)14941461
EQ-Bench 4EQ-Bench1294–
Arena Creative WritingArena (LMArena)14821438
Arena Multi-turnArena (LMArena)15141468
Arena Longer QueriesArena (LMArena)15081474
Multimodal
Arena VisionArena (LMArena)1317–
MMMU-ProArtificial Analysis78.8%†–
MMMU-Pro (Vals)Vals AI85.5%–
MortgageTaxVals AI70.3%–
SAGEVals AI56.1%–
Blueprint-Bench 2Epoch AI Benchmarking Hub24.5%–
Long context
AA-LCRArtificial Analysis78.7%†80.7%
Arena DocumentArena (LMArena)1494–
Composite
Epoch Capabilities IndexEpoch AI Benchmarking Hub156.8149.4
LiveBench 1.0LiveBench76.5%†–
AA Intelligence Index v4.3.2Artificial Analysis40.7†28.4
Vals RSI IndexVals AI19.1%†–
Vals Multimodal IndexVals AI67.4%–

Data as of 2026-10-26. Best configuration of each model; every score links to its source on the model pages.

What changed

  • Claude Opus 4.7: 1 score revision.
  • Claude Opus 4.7: 1 new result (Harvey's Legal Agent Benchmark v1).
  • Claude Opus 4.7: 1 score revision.
  • Claude Opus 4.7: 1 score revision.
  • Qwen3.6 Max (max): BenchLeader Index 62.8 → 62.3.
  • Claude Opus 4.7: 1 score revision.
  • Claude Opus 4.7: BenchLeader Index 64.0 → 63.5.
  • Claude Opus 4.7: 1 new result (SWE Atlas: Codebase QnA).
  • Claude Opus 4.7: BenchLeader Index 64.5 → 64.0.
  • Claude Opus 4.7: 1 new result (SWE Atlas: Test Writing).

Rankings: Best LLM overall · Best LLM for agents · Best cheap LLMs

Claude Opus 4.7 vs Qwen3.6 Max: questions

Is Claude Opus 4.7 better than Qwen3.6 Max?
Claude Opus 4.7 and Qwen3.6 Max (max) are neck and neck on quality (63.2 vs 62.3, within the ±4.0 margin). The BenchLeader Index combines every independent quality benchmark; the two are within the margin of error as of 2026-10-26, so check the category scores for your use.
Which is cheaper, Claude Opus 4.7 or Qwen3.6 Max, and will that change?
Qwen3.6 Max is cheaper today: $2.96 against $10.00 per million tokens, blended at three input tokens per output token. No price change has been announced for either.
Is Claude Opus 4.7 better than Qwen3.6 Max for coding?
Claude Opus 4.7 scores higher in coding (61 vs 57 on the category index, where 50 is average). On Arena WebDev, the top shared coding benchmark, it is 1555 for Claude Opus 4.7 against 1482 for Qwen3.6 Max.
Is Claude Opus 4.7 better than Qwen3.6 Max for agents and tool use?
Qwen3.6 Max scores higher in agentic tasks (72 vs 65 on the category index, where 50 is average).
Which is better for long documents?
Claude Opus 4.7 accepts more context: 1,000,000 against 245,800 tokens for Qwen3.6 Max.
Do independent benchmarks agree on Claude Opus 4.7 vs Qwen3.6 Max?
Across the 15 benchmarks both have, Claude Opus 4.7 leads 13, Qwen3.6 Max leads 2. Every overall index that rates both puts Claude Opus 4.7 ahead.
When was Qwen3.6 Max released?
Qwen3.6 Max was released on 20 Apr 2026.