BenchLeader
Menu
Best LLMsChoose a modelLeaderboardsModelsBenchmarksProvidersReleases9

GPT-5.4 vs Qwen3.6 Max

Full profiles: GPT-5.4 (OpenAI) and Qwen3.6 Max (Alibaba): every benchmark, price and rank.

Updated · Data from 11 sources

As of 27 Oct 2026, GPT-5.4 (xhigh) and Qwen3.6 Max (max) are neck and neck on quality (64.6 vs 62.2 on the BenchLeader Index, inside the ±4.0 margin of error), and Qwen3.6 Max (max) costs 1.9× less, which makes it the better buy for most workloads. Pick GPT-5.4 for reasoning (69 vs 58); pick Qwen3.6 Max for agents & tools (72 vs 64).

GPT-5.4 benchmarks xhigh OpenAI 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
CodingGPT-5.465 vs 57 category score; SWE-bench (Vals) 78.2% vs 72.8%High
Agents & tool useQwen3.6 Max72 vs 64 category score; Terminal-Bench Hard 43.9% vs 57.6%High
MathsEither65 vs 65 category score; OTIS Mock AIME 95.3% vs 91.1%Either
ReasoningGPT-5.469 vs 58 category score; GPQA Diamond 93.3% vs 87.4%High
KnowledgeQwen3.6 Max62 vs 60 category score; SimpleQA Verified 52.0% vs 45.1%Medium
Long documentsEither66 vs 65 category score; AA-LCR 82.0% vs 80.7%Either
Following instructionsQwen3.6 Max75 vs 72 category score; IFBench 76.6% vs 74.0%Medium
Chat quality (human preference)–Only Qwen3.6 Max has an Arena rating (65 category score)Not comparable
Lowest costQwen3.6 Max$2.92 vs $5.63 per 1M blended; agentic coding session $0.109 vs $0.210–
High volume on a budgetNeither flagshipA smaller model is usually the better fit: Cheapest good LLMs, Fastest LLMs–

Key differences

  • GPT-5.4 (xhigh) and Qwen3.6 Max (max) are neck and neck on quality (64.6 vs 62.2, within the ±4.0 margin).
  • GPT-5.4 (xhigh) is stronger in coding, composite, long context, maths, multimodal, reasoning.
  • Qwen3.6 Max (max) is stronger in agents & tools, instruction following, knowledge, human preference.
  • Qwen3.6 Max (max) is 1.9× cheaper ($2.92 vs $5.63 per 1M blended).

At a glance

SpecGPT-5.4Qwen3.6 Max
DeveloperOpenAIAlibaba
Released5 Mar 202620 Apr 2026
Context window (tokens)1,050,000262,144
Compared atGPT-5.4 (xhigh)Qwen3.6 Max (max)
Open weightsNoNo
Price in / out / cached input per 1M$2.50 / $15.00 / $0.250$1.30 / $7.80 / $0.130

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.

CategoryGPT-5.4 (xhigh)Qwen3.6 Max (max)Leader
Reasoning68.958.1GPT-5.4
Coding64.656.6GPT-5.4
Agents & tools63.672.1Qwen3.6 Max
Maths65.264.9level
Knowledge59.762.1Qwen3.6 Max
Instruction following72.374.6Qwen3.6 Max
Human preference–64.8not comparable
Multimodal61.1–not comparable
Long context65.965.2level
Composite72.660.9GPT-5.4

Where sources disagree

Across the 18 benchmarks both have, GPT-5.4 leads 12, Qwen3.6 Max leads 6. Every overall index that rates both puts GPT-5.4 ahead.

Overall indexGPT-5.4 (xhigh)Qwen3.6 Max (max)Leader
BenchLeader Index64.662.2neck and neck
AA Intelligence Index v4.3.239.028.4GPT-5.4
Arena Text1475†1461GPT-5.4

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.

MeasureGPT-5.4 (xhigh)Qwen3.6 Max (max)
Blended price per 1M tokens$5.63$2.92
Input / output per 1M$2.50 / $15.00$1.30 / $7.80
Cached input per 1M$0.250$0.130
Price per task: chat reply$0.0055$0.0029
Price per task: summarize a 30-page report$0.039$0.020
Price per task: code edit$0.037$0.019
Price per task: agentic coding session$0.210$0.109
Price per task: structured extraction$0.0080$0.0042
Output speed117 tok/s–
Time to first answer148.3 s–
Time to a complete answer152.6 s–

Each model is compared at its best-ranked reasoning-effort setting. GPT-5.4 (xhigh) scores 64.6 and takes 148 s to its first answer token; GPT-5.4 (low) scores 59.6 in 2 s, GPT-5.4 (high) scores 58.9 in 1 s. Qwen3.6 Max (max) scores 62.2. 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

113 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.

MetricPublisherGPT-5.4 (xhigh)Qwen3.6 Max (max)
BenchLeader IndexBenchLeader64.662.2
Agents & tools scoreBenchLeader63.672.1
Coding scoreBenchLeader64.656.6
Composite scoreBenchLeader72.660.9
Instruction following scoreBenchLeader72.374.6
Knowledge scoreBenchLeader59.762.1
Long context scoreBenchLeader65.965.2
Maths scoreBenchLeader65.264.9
Multimodal scoreBenchLeader61.1–
Reasoning scoreBenchLeader68.958.1
Human preference scoreBenchLeader–64.8
Blended price $/MBenchLeader$5.63$2.92
Output speedBenchLeader117 tok/s–
Time to first answerBenchLeader148.3 s–
Context windowBenchLeader1,050,000 tokens262,144 tokens
Reasoning
GPQA DiamondEpoch AI Benchmarking Hub93.3%87.4%
Humanity's Last ExamScale AI / CAIS36.2%–
SimpleBenchSimpleBench–63.0%
Arena Hard PromptsArena (LMArena)1487†1483
LiveBench 1.0 ReasoningLiveBench88.1%–
GPQA Diamond (AA)Artificial Analysis92.0%88.8%
Humanity's Last Exam (AA)Artificial Analysis43.7%30.8%
GPQA Diamond (Vals)Vals AI91.7%–
Kagi LLM BenchmarkKagi LLM Benchmark63.8%†–
ARC-AGI-1ARC Prize93.7%–
ARC-AGI-2ARC Prize74.0%–
ARC-AGI-3ARC Prize0.2%†–
CritPtArtificial Analysis23.4%3.7%
Chess PuzzlesEpoch AI Benchmarking Hub48.0%–
Mystery Game PuzzlesEpoch AI Benchmarking Hub37.0%19.0%†
CL-benchEpoch AI Benchmarking Hub27.9%–
CL-bench LifeEpoch AI Benchmarking Hub21.7%–
LMCAEpoch AI Benchmarking Hub52.0%42.5%
DTBenchEpoch AI Benchmarking Hub94.4%87.2%
ForecastBenchForecastBench63.1%†–
Coding
SWE-bench Verified (Epoch)Epoch AI Benchmarking Hub76.9%†76.7%
SciCodeSciCode56.6%–
WeirdMLWeirdML77.7%–
GSO-BenchGSO-Bench31.4%–
Arena CodingArena (LMArena)1511†1512
Show all 113 benchmarks
MetricPublisherGPT-5.4 (xhigh)Qwen3.6 Max (max)
Coding
Arena WebDevArena (LMArena)1401†1482
LiveBench 1.0 CodingLiveBench77.5%–
LiveCodeBenchVals AI84.1%–
SWE-bench (Vals)Vals AI78.2%72.8%
SWE-Bench ProScale AI SEAL59.1%–
Code MigrationVals AI35.0%–
ProgramBenchVals AI0.0%–
Vibe Code Bench v1.1Vals AI67.4%–
SWE-Bench Pro (private)Scale AI SEAL59.1%–
SWE Atlas: Codebase QnAScale AI SEAL40.8%–
SWE Atlas: RefactoringScale AI SEAL44.3%–
SWE Atlas: Test WritingScale AI SEAL44.4%–
MirrorCodeEpoch AI Benchmarking Hub15.6%†–
DeepSWE v1.1Epoch AI Benchmarking Hub51.8%–
ALE-BenchEpoch AI Benchmarking Hub1607†–
AlgoTuneAlgoTune1.9†–
Terminal-Bench 2.1 (AA)Artificial Analysis78.3%–
Agents & tools
Terminal-BenchTerminal-Bench81.8%†–
APEX-AgentsMercor52.4%†–
LiveBench Agentic CodingLiveBench53.8%–
Terminal-Bench HardArtificial Analysis57.6%43.9%
τ²-Bench Telecom (AA)Artificial Analysis87.1%95.9%
MCP AtlasScale AI SEAL70.6%–
HiL-BenchScale AI SEAL9.7%†–
τ³-bench Bankingτ²-bench39.4%–
τ³-Banking (AA)Artificial Analysis39.6%–
APEX-Agents (AA)Artificial Analysis33.3%–
Harvey's Legal Agent Benchmark v1Vals AI0.0%–
SkillsBenchVals AI51.7%–
Terminal-Bench 2.0 (Vals)Vals AI58.4%†51.7%
DeepResearch BenchDeepResearch Bench35.1%†–
PostTrainBenchEpoch AI Benchmarking Hub19.0%†–
METR Time HorizonsMETR74.3%–
GBAEvalEpoch AI Benchmarking Hub45.1%†–
Vending-Bench 2Andon Labs6144.2†4254.2
DeepSearchQAGoogle63.7%†–
Maths
FrontierMath Tiers 1–3Epoch AI Benchmarking Hub78.6%–
FrontierMath Tier 4Epoch AI Benchmarking Hub49.0%–
OTIS Mock AIMEEpoch AI Benchmarking Hub95.3%91.1%
ProofBenchVals AI56.0%–
LiveBench MathematicsLiveBench94.2%–
AIME (Vals)Vals AI96.7%–
AIME 2026MathArena99.2%–
HMMT February 2026MathArena97.7%–
USAMO 2026MathArena95.2%–
MathArena ApexMathArena54.2%–
Arena MathsArena (LMArena)1460†1476
Knowledge
SimpleQA VerifiedEpoch AI Benchmarking Hub45.1%52.0%
LiveBench 1.0 Data AnalysisLiveBench79.3%–
AA-OmniscienceArtificial Analysis5.89.2
MMLU-ProVals AI87.5%–
LegalBenchVals AI86.0%–
CorpFinVals AI65.3%66.5%
TaxEvalVals AI74.0%–
MedQAVals AI96.1%–
PRBench FinanceScale AI SEAL45.6%†–
PRBench LegalScale AI SEAL44.4%†–
MultiNRCScale AI SEAL58.3%†–
CaseLaw v2Vals AI63.8%47.9%
MedCodeVals AI41.3%–
MedScribeVals AI77.5%–
EBR-benchEpoch AI Benchmarking Hub25.4%–
AA-Omniscience: accuracyArtificial Analysis50.9%37.9%
AA-Omniscience: non-hallucinationArtificial Analysis8.3%53.8%
Instruction following
LiveBench LanguageLiveBench82.6%–
LiveBench Instruction FollowingLiveBench70.2%–
IFBenchArtificial Analysis74.0%76.6%
Arena Instruction FollowingArena (LMArena)1460†1452
Human preference
Arena TextArena (LMArena)1465†1461
EQ-Bench 4EQ-Bench1260†–
Arena Creative WritingArena (LMArena)1435†1438
Arena Multi-turnArena (LMArena)1480†1468
Arena Longer QueriesArena (LMArena)1477†1474
Multimodal
Arena VisionArena (LMArena)1292†–
MMMU-ProArtificial Analysis78.4%–
VISTAScale AI SEAL50.9%†–
MMMU-Pro (Vals)Vals AI87.5%–
MortgageTaxVals AI68.3%–
SAGEVals AI43.3%–
VTBScale AI SEAL29.2%†–
Blueprint-Bench 2Epoch AI Benchmarking Hub27.1%†–
Long context
AA-LCRArtificial Analysis82.0%80.7%
Arena DocumentArena (LMArena)1470†–
Composite
Epoch Capabilities IndexEpoch AI Benchmarking Hub157.3†149.4
LiveBench 1.0LiveBench78.0%–
AA Intelligence Index v4.3.2Artificial Analysis39.028.4
Vals RSI IndexVals AI15.5%–
Safety & honesty
MASKScale AI SEAL89.7%†–

Data as of 2026-10-27. Best configuration of each model; every score links to its source on the model pages.

What changed

  • GPT-5.4 (xhigh): 2 new results (Harvey's Legal Agent Benchmark v1, Vals RSI Index).
  • GPT-5.4 (xhigh): 2 new results (USAMO 2026, τ³-bench Banking).
  • Qwen3.6 Max (max): BenchLeader Index 62.8 → 62.3.
  • GPT-5.4 (xhigh): 1 score revision.
  • GPT-5.4 (xhigh): 1 new result (SWE Atlas: Codebase QnA).
  • GPT-5.4 (xhigh): 1 score revision.
  • GPT-5.4 (xhigh): BenchLeader Index 64.9 → 64.4.
  • GPT-5.4 (xhigh): 1 new result (SWE Atlas: Test Writing).
  • Qwen3.6 Max (max): BenchLeader Index 63.3 → 62.8.
  • GPT-5.4 (xhigh): 1 new result (SWE Atlas: Codebase QnA).

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

GPT-5.4 vs Qwen3.6 Max: questions

Is GPT-5.4 better than Qwen3.6 Max?
GPT-5.4 (xhigh) and Qwen3.6 Max (max) are neck and neck on quality (64.6 vs 62.2, within the ±4.0 margin). The BenchLeader Index combines every independent quality benchmark; the two are within the margin of error as of 2026-10-27, so check the category scores for your use.
Which is cheaper, GPT-5.4 or Qwen3.6 Max, and will that change?
Qwen3.6 Max is cheaper today: $2.92 against $5.63 per million tokens, blended at three input tokens per output token. No price change has been announced for either.
Is GPT-5.4 better than Qwen3.6 Max for coding?
GPT-5.4 scores higher in coding (65 vs 57 on the category index, where 50 is average).
Is GPT-5.4 better than Qwen3.6 Max for agents and tool use?
Qwen3.6 Max scores higher in agentic tasks (72 vs 64 on the category index, where 50 is average). On Terminal-Bench Hard, the top shared agents & tools benchmark, it is 57.6% for GPT-5.4 against 43.9% for Qwen3.6 Max.
Which is better for long documents?
They are level on long-context benchmarks (66 vs 65). GPT-5.4 accepts more context: 1,050,000 against 262,144 tokens for Qwen3.6 Max.
Do independent benchmarks agree on GPT-5.4 vs Qwen3.6 Max?
Across the 18 benchmarks both have, GPT-5.4 leads 12, Qwen3.6 Max leads 6. Every overall index that rates both puts GPT-5.4 ahead.
When was Qwen3.6 Max released?
Qwen3.6 Max was released on 20 Apr 2026.