BenchLeader

GPT-5.2 vs Qwen3.8 Max

Full profiles: GPT-5.2 (OpenAI) and Qwen3.8 Max (Alibaba): every benchmark, price and rank.

Updated · Data from 14 sources

As of 16 Nov 2026, GPT-5.2 and Qwen3.8 Max are neck and neck on quality (59.9 vs 58.7 on the BenchLeader Index, inside the ±5.2 margin), and Qwen3.8 Max is 1.6× cheaper at $3.00 vs $4.81 per 1M tokens blended. Choose GPT-5.2 for reasoning (60 vs 52), coding (63 vs 62), and long context (66 vs 65), where it also streams faster; choose Qwen3.8 Max for maths (65 vs 57), agents & tools (59 vs 54), and knowledge (58 vs 56).

Key facts (as of 16 Nov 2026)

  • GPT-5.2 scores 59.9 on the BenchLeader Index and Qwen3.8 Max scores 58.7 (BenchLeader, 16 Nov 2026).
  • GPT-5.2 costs $4.81 per 1M tokens blended (list price, 16 Nov 2026).
  • Qwen3.8 Max costs $3.00 per 1M tokens blended (list price, 16 Nov 2026).
  • GPT-5.2 streams faster than Qwen3.8 Max (Artificial Analysis, 7-day median, 16 Nov 2026).
  • Of the 20 benchmarks both models have, Qwen3.8 Max leads 15 (BenchLeader, 16 Nov 2026).

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-5.263 vs 62 category score; LiveCodeBench 85.4% vs 87.8%Medium
Agents & tool useQwen3.8 Max59 vs 54 category scoreMedium
MathsQwen3.8 Max65 vs 57 category score; ProofBench 58.0% vs 15.0%High
ReasoningGPT-5.260 vs 52 category score; GPQA Diamond (AA) 90.3% vs 92.8%High
KnowledgeQwen3.8 Max58 vs 56 category score; MMLU-Pro 88.6% vs 86.2%Medium
Long documentsGPT-5.266 vs 65 category score; AA-LCR v1.1 82.7% vs 80.3%Medium
Chat quality (human preference)–Only Qwen3.8 Max has an Arena rating (68 category score)Not comparable
Lowest costQwen3.8 Max$3.00 vs $4.81 per 1M blended; agentic coding session $0.144 vs $0.161–
Fastest first answerEither112.5 s (GPT-5.2 (xhigh)) vs 119.6 s (Qwen3.8 Max (max)) to the first answer token–
High volume on a budgetNeither flagshipA smaller model is usually the better fit: Cheapest good LLMs, Fastest LLMs–

Key differences

  • GPT-5.2 (xhigh) and Qwen3.8 Max (max) are neck and neck on quality (59.9 vs 58.7, within the ±5.2 margin).
  • GPT-5.2 (xhigh) is stronger in coding, long context, reasoning.
  • Qwen3.8 Max (max) is stronger in agents & tools, knowledge, maths.
  • Qwen3.8 Max (max) is 1.6× cheaper ($3.00 vs $4.81 per 1M blended).

At a glance

SpecGPT-5.2Qwen3.8 Max
DeveloperOpenAIAlibaba
Released11 Dec 20253 Aug 2026
Context window (tokens)400,0001,000,000
Compared atGPT-5.2 (xhigh)Qwen3.8 Max (max)
Open weightsNoNo
Price in / out / cached input per 1M$1.75 / $14.00 / $0.175$2.00 / $6.00 / $0.250

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-5.2 (xhigh)Qwen3.8 Max (max)Leader
Reasoning60.451.5GPT-5.2
Coding63.061.6GPT-5.2
Agents & tools53.858.5Qwen3.8 Max
Maths56.565.2Qwen3.8 Max
Knowledge56.158.0Qwen3.8 Max
Instruction following73.7–not comparable
Human preference–67.5not comparable
Multimodal–65.7not comparable
Long context66.064.8GPT-5.2
Composite–45.0not comparable

On one chart

Each domain on the index's own scale, where 50 is an average model. Click a domain to see the benchmarks behind it. Pick up to 5 models from the list.

2550 avg75ReasoningCodingAgents & toolsMathsKnowledgeLong contextGPT-5.2, Reasoning: 60.4GPT-5.2, Coding: 63.0GPT-5.2, Agents & tools: 53.8GPT-5.2, Maths: 56.5GPT-5.2, Knowledge: 56.1GPT-5.2, Long context: 66.0Qwen3.8 Max, Reasoning: 51.5Qwen3.8 Max, Coding: 61.6Qwen3.8 Max, Agents & tools: 58.5Qwen3.8 Max, Maths: 65.2Qwen3.8 Max, Knowledge: 58.0Qwen3.8 Max, Long context: 64.8

Not drawn, measured for fewer than two of these models: Instruction following, Multimodal, Human preference.

Model, click to add or removeIndex

Where sources disagree

Across the 20 benchmarks both have, GPT-5.2 leads 5, Qwen3.8 Max leads 15. Every overall index that rates both puts Qwen3.8 Max ahead.

Overall indexGPT-5.2 (xhigh)Qwen3.8 Max (max)Leader
BenchLeader Index59.958.7neck and neck
Arena Text1437†1484Qwen3.8 Max

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. Cost, tokens and time per benchmark task are measured: Artificial Analysis’ averages over its whole Intelligence Index run.

MeasureGPT-5.2 (xhigh)Qwen3.8 Max (max)
Blended price per 1M tokens$4.81$3.00
Input / output per 1M$1.75 / $14.00$2.00 / $6.00
Cached input per 1M$0.175$0.250
Price per task: chat reply$0.0049$0.0026
Price per task: summarize a 30-page report$0.029$0.028
Price per task: code edit$0.032$0.021
Price per task: agentic coding session$0.161$0.144
Price per task: structured extraction$0.0063$0.0052
Agentic coding session, input mostly cached$0.090$0.065
Cost per benchmark task (AA Index)–$10.02
Output tokens per task (AA Index)–171k
Of which reasoning–120k
Time per task (AA Index)–48 min
Output speed76 tok/s59 tok/s
Time to first answer112.5 s119.6 s
Time to a complete answer, last measured 13 Nov 2026154.4 s294.1 s

Each model is compared at its best-ranked reasoning-effort setting. GPT-5.2 (xhigh) scores 59.9 and takes 112 s to its first answer token; GPT-5.2 scores 56.7 in 2 s, GPT-5.2 (medium) scores 56.2 in 7 s. Qwen3.8 Max (max) scores 58.7 and takes 120 s to its first answer token; Qwen3.8 Max (xhigh) scores 56.6 in 3 s. 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

188 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-5.2 (xhigh)Qwen3.8 Max (max)
BenchLeader IndexBenchLeader59.958.7
Agents & tools scoreBenchLeader53.858.5
Coding scoreBenchLeader63.061.6
Instruction following scoreBenchLeader73.7–
Knowledge scoreBenchLeader56.158.0
Long context scoreBenchLeader66.064.8
Maths scoreBenchLeader56.565.2
Reasoning scoreBenchLeader60.451.5
Composite scoreBenchLeader–45.0
Human preference scoreBenchLeader–67.5
Multimodal scoreBenchLeader–65.7
Blended price $/MBenchLeader$4.81$3.00
Output speedBenchLeader76 tok/s59 tok/s
Time to first answerBenchLeader112.5 s119.6 s
Context windowBenchLeader400,000 tokens1,000,000 tokens
Reasoning
GPQA DiamondEpoch AI Benchmarking Hub91.4%92.7%†
Humanity's Last ExamScale AI / CAIS27.8%†–
SimpleBenchSimpleBench45.8%†–
Arena Hard PromptsArena (LMArena)1497†1507
LiveBench 2.0 ReasoningLiveBench–88.2%
LiveBench 1.0 ReasoningLiveBench83.2%†88.2%
GPQA Diamond (AA)Artificial Analysis90.3%92.8%
Humanity's Last Exam (AA)Artificial Analysis37.7%43.1%
HLE-Diamond (AA)Artificial Analysis–24.0%
GPQA Diamond (Vals)Vals AI91.7%93.7%
Kagi LLM BenchmarkKagi LLM Benchmark73.3%†–
ARC-AGI-1ARC Prize86.2%–
ARC-AGI-2ARC Prize52.9%–
CritPtArtificial Analysis11.6%20.0%
CritPt v2 hard subset (AA)Artificial Analysis–29.5%
MysteryMechanismVals AI–23.9%
SciPredictScale AI SEAL20.6%†–
Arena ExpertArena (LMArena)1491†1514
Chess PuzzlesEpoch AI Benchmarking Hub48.0%46.0%†
Mystery Game PuzzlesEpoch AI Benchmarking Hub22.0%†38.0%†
CL-benchEpoch AI Benchmarking Hub18.2%†–
LMCAEpoch AI Benchmarking Hub43.9%46.2%†
DTBenchEpoch AI Benchmarking Hub90.9%92.0%†
ForecastBenchForecastBench63.8%†–
Coding
SWE-bench Verified (Epoch)Epoch AI Benchmarking Hub73.8%†–
Show all 188 benchmarks
MetricPublisherGPT-5.2 (xhigh)Qwen3.8 Max (max)
Coding
SciCodeSciCode–53.2%
WeirdMLWeirdML72.2%–
GSO-BenchGSO-Bench27.4%†–
Arena CodingArena (LMArena)1515†1523
Arena WebDevArena (LMArena)1421†1676
LiveBench 2.0 CodingLiveBench–60.9%
LiveBench 1.0 CodingLiveBench76.1%†72.9%
LiveCodeBenchVals AI85.4%87.8%
IOIVals AI–68.9%
SWE-bench (Vals)Vals AI75.8%85.6%
SWE-Bench ProScale AI SEAL29.9%†–
SWE-bench Verified (bash only)SWE-bench69.0%†–
SWE-bench Multilingual (bash only)SWE-bench66.7%†–
SWE-bench Verified (any scaffold)SWE-bench72.8%†–
Code MigrationVals AI–24.0%
ProgramBenchVals AI–0.0%
SWE-bench MultimodalVals AI–42.5%
Vibe Code Bench 1-100Vals AI–12.8%
Vibe Code Bench v1.1Vals AI53.5%64.7%
SWE-Bench Pro (private)Scale AI SEAL23.8%†–
SWE Atlas v1.1Scale AI SEAL–64.4%
SWE Atlas v1.1: Codebase QnAScale AI SEAL–81.5%
SWE Atlas v1.1: Test WritingScale AI SEAL–58.9%
SWE Atlas v1.1: RefactoringScale AI SEAL–52.9%
MirrorCodeEpoch AI Benchmarking Hub32.2%52.2%†
DeepSWE v1.1Epoch AI Benchmarking Hub–57.5%†
ALE-BenchEpoch AI Benchmarking Hub1249.8†1679.1†
AlgoTuneAlgoTune2.0†–
FrontierSWEEpoch AI Benchmarking Hub–17.8%†
WeirdML v3WeirdML5.0%–
Terminal-Bench 4.0 (AA)Artificial Analysis–38.9%
Terminal-Bench 2.1 (AA)Artificial Analysis–88.8%
Startrise FrontendStartrise–71.3%
GameBench 2 OverallDuelLab GameBench 2–54.2†
GameBench 2 by reasoning settingDuelLab GameBench 2–69.6†
Design Arena WebsiteDesign Arena11861269
FrontierCode 1.1 (AA)Artificial Analysis–43.2%
FrontierCode 2.0 (AA)Artificial Analysis–23.4%
LiveCodeBench (AA)Artificial Analysis88.9%–
Agents & tools
Terminal-BenchTerminal-Bench64.9%†27.0%
GDPvalOpenAI49.7%†–
Remote Labor IndexScale AI / CAIS2.1%†–
APEX-AgentsMercor–63.3%
Arena AgentArena (LMArena)–0.1
LiveBench Long AgenticLiveBench–26.6%
LiveBench BrowserLiveBench–49.4%
LiveBench Agentic CodingLiveBench50.3%†64.7%
Terminal-Bench HardArtificial Analysis47.0%–
τ²-Bench Telecom (AA)Artificial Analysis84.8%–
Terminal-Bench 2.1 (Vals)Vals AI–67.4%
MCP AtlasScale AI SEAL67.6%–
BFCL OverallBerkeley Function Calling Leaderboard55.9%†–
τ²-benchτ²-bench84.8%†–
τ³-bench Bankingτ²-bench32.2%†55.1%†
GDPval-AA v2.1, normalised scoreArtificial Analysis–58.3%
GDPval-AA v2.1Artificial Analysis–1674
τ³-Banking (AA)Artificial Analysis11.1%†51.3%
ITBench SRE (AA)Artificial Analysis–40.3%
Analyst Agent (AA)Artificial Analysis–45.0%
APEX-Agents (AA)Artificial Analysis–42.4%
CyberBenchVals AI–28.6%
Finance Agent v2Vals AI–50.6%
Harvey's Legal Agent Benchmark v1Vals AI–10.4%
Legal Research BenchVals AI–47.6%
Poker AgentVals AI1131.8%†–
SkillsBenchVals AI–42.0%
Tax Agent BenchVals AI–32.1%
Terminal-Bench 2.0 (Vals)Vals AI51.7%†–
Terminal-Bench 4.0 (Vals)Vals AI–34.3%
Terminal-Bench ScienceVals AI–12.9%
Arena SearchArena (LMArena)1161†1227
Arena Agent Task CompletionArena (LMArena)–4.5
DeepResearch BenchDeepResearch Bench41.1%†–
METR Time HorizonsMETR75.3%†–
Vending-Bench 2Andon Labs3591.3†–
AutomationBenchArtificial Analysis–56.2%
GDP.pdf (AA)Artificial Analysis–22.8%
EnterpriseOps-GymArtificial Analysis–47.6%
AA-Briefcase v1.1Artificial Analysis–1604
GDPxGDPx by SemiAnalysis–3.5%
Design Arena AgenticDesign Arena1065†1264
Terminal-Bench-Science 0.1 (AA)Artificial Analysis–11.9%
MKA (AA)Artificial Analysis–37.4%
FACTS SearchGoogle–85.0%
Maths
FrontierMath Tiers 1–3Epoch AI Benchmarking Hub67.4%74.7%†
FrontierMath Tier 4Epoch AI Benchmarking Hub31.7%46.3%†
OTIS Mock AIMEEpoch AI Benchmarking Hub96.1%100.0%†
ProofBenchVals AI15.0%58.0%
LiveBench MathematicsLiveBench93.2%†91.3%
AIME (Vals)Vals AI96.9%–
MGSMVals AI94.0%–
AIME 2026MathArena98.3%†–
HMMT February 2026MathArena97.0%†–
MathArena ApexMathArena13.5%†–
Arena MathsArena (LMArena)1453†1505
AIME 2025 (AA)Artificial Analysis99.0%–
Knowledge
SimpleQA VerifiedEpoch AI Benchmarking Hub37.1%47.3%†
LiveBench 2.0 Data AnalysisLiveBench–78.4%
LiveBench ResearchLiveBench–72.8%
LiveBench 1.0 Data AnalysisLiveBench78.2%†78.4%
AA-Omniscience v1.1Artificial Analysis–10.5
MMLU-ProVals AI86.2%88.6%
LegalBenchVals AI82.8%83.6%
CorpFinVals AI65.9%65.8%
TaxEvalVals AI75.8%75.5%
MedQAVals AI94.1%–
MultiNRCScale AI SEAL42.2%†–
CaseLaw v2Vals AI66.0%–
Excel Modeling BenchmarkVals AI–60.1%
Excel Modeling Benchmark v1.1Vals AI–66.0%
MedCodeVals AI49.8%40.7%
MedScribeVals AI84.4%85.0%
Public Benefits Bench v1.1Vals AI–67.1%
Access to Justice BenchmarkVals AI–76.4%
Financial Coaching BenchVals AI–75.6%
Immigration BenchVals AI–70.1%
DrugDiscoveryBenchScale AI SEAL29.3%–
EBR-benchEpoch AI Benchmarking Hub23.0%–
GDP.pdfEpoch AI Benchmarking Hub–23.2%†
AA-Omniscience v1.1: accuracyArtificial Analysis–29.5%
AA-Omniscience v1.1: non-hallucinationArtificial Analysis–72.9%
Global-MMLU-Lite (AA)Artificial Analysis89.8%†–
FACTS ParametricGoogle–30.8%
Instruction following
LiveBench LanguageLiveBench79.8%†79.7%
LiveBench Instruction FollowingLiveBench61.8%†74.1%
IFBenchArtificial Analysis75.4%–
TutorBenchScale AI SEAL53.5%†–
Arena Instruction FollowingArena (LMArena)1457†1478
Human preference
Arena TextArena (LMArena)1476†1484
EQ-Bench Creative Writing v3EQ-Bench1703†–
Arena Creative WritingArena (LMArena)1432†1472
Arena Multi-turnArena (LMArena)1495†1495
Arena Longer QueriesArena (LMArena)1472†1496
Multimodal
Arena VisionArena (LMArena)1268†1312
LiveBench DesignLiveBench–76.6%
MMMU-ProArtificial Analysis74.6%†82.8%
VISTAScale AI SEAL46.6%†–
MMMU-Pro (Vals)Vals AI86.7%88.0%
MortgageTaxVals AI67.1%64.0%
SAGEVals AI–81.8%
Humanity's Sixth SenseScale AI SEAL–33.0%
VPCTEpoch AI Benchmarking Hub84.0%–
Furniture AssemblyEpoch AI Benchmarking Hub38.3%20.0%†
VIALS-AAHandshake AI–31.7%
Code2Video-AAArtificial Analysis–1021
AA-3DArtificial Analysis–954
FACTS MultimodalGoogle–51.5%
Long context
AA-LCR v1.1Artificial Analysis82.7%80.3%
Arena DocumentArena (LMArena)1425†–
MLCRArtificial Analysis–26.1%
FACTS GroundingGoogle–73.8%
Composite
Epoch Capabilities IndexEpoch AI Benchmarking Hub154.5†157.6
LiveBench 2.0LiveBench–60.9%
LiveBench 1.0LiveBench74.6%†78.5%
AA Intelligence Index v5.0.1Artificial Analysis–33.6
AA Intelligence Index v4.3.2Artificial Analysis30.445.4
AA Progress Index v1Artificial Analysis295.5323.7
Vals RSI IndexVals AI15.8%21.8%
Vals Multimodal IndexVals AI–65.4%
Safety & honesty
FORTRESSScale AI SEAL17.5%†–
MASKScale AI SEAL86.7%†–
PropensityBenchScale AI SEAL34.4%†–
DistressBenchScale AI SEAL–77.6%

Data as of 2026-11-16. Best configuration of each model; every score links to its source on the model pages.

DuelLab GameBench 2, release gb2-20261111-7b598bb13de4 (https://benchmarks.duellab.org/releases/gb2-20261111-7b598bb13de4/), CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). Scores mapped to BenchLeader model names; not endorsed by DuelLab.

What changed

  • Qwen3.8 Max (max): 1 score revision.
  • GPT-5.2 (xhigh): 1 new result (AIME 2025 (AA)).
  • Qwen3.8 Max (max): 1 score revision.
  • GPT-5.2 (xhigh): 1 new result (MirrorCode).
  • Qwen3.8 Max (max): 6 score revisions.
  • Qwen3.8 Max (max): 1 score revision.
  • Qwen3.8 Max (max): 1 new result (FACTS Search); 1 score revision.
  • Qwen3.8 Max (max): 1 score revision; cached input price moved from $0.22/M to $0.25/M (13% up).
  • Qwen3.8 Max (max): cached input price moved from $0.25/M to $0.22/M (11% down).
  • Qwen3.8 Max (max): 1 score revision; cached input price moved from $0.22/M to $0.25/M (13% up).

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

GPT-5.2 vs Qwen3.8 Max: questions

Is GPT-5.2 better than Qwen3.8 Max?
GPT-5.2 (xhigh) and Qwen3.8 Max (max) are neck and neck on quality (59.9 vs 58.7, within the ±5.2 margin). The BenchLeader Index combines every independent quality benchmark; the two are within the margin of error as of 16 Nov 2026, so check the category scores for your use.
GPT-5.2 vs Qwen3.8 Max: which is cheaper, and will that change?
Qwen3.8 Max is cheaper today: $3.00 against $4.81 per million tokens, blended at three input tokens per output token. No price change has been announced for either.
Is GPT-5.2 or Qwen3.8 Max better for coding?
GPT-5.2 scores higher in coding (63 vs 62 on the category index, where 50 is average). On LiveCodeBench, the top shared coding benchmark, it is 85.4% for GPT-5.2 against 87.8% for Qwen3.8 Max.
Which is faster, GPT-5.2 or Qwen3.8 Max?
GPT-5.2 streams faster (76 vs 59 output tokens per second, Artificial Analysis 7-day median).
Is GPT-5.2 or Qwen3.8 Max better for agents and tool use?
Qwen3.8 Max scores higher in agentic tasks (59 vs 54 on the category index, where 50 is average).
More questions (4)
Which is better for long documents?
GPT-5.2 scores higher on long-context benchmarks (66 vs 65). Qwen3.8 Max accepts more context: 1,000,000 against 400,000 tokens for GPT-5.2.
Which has lower latency?
GPT-5.2 answers first: 112.5 s against 119.6 s to the first answer token, as Artificial Analysis measures it. Reasoning effort drives most of this: a higher setting thinks longer before it answers.
Do independent benchmarks agree on GPT-5.2 vs Qwen3.8 Max?
Across the 20 benchmarks both have, GPT-5.2 leads 5, Qwen3.8 Max leads 15. Every overall index that rates both puts Qwen3.8 Max ahead.
When was Qwen3.8 Max released?
Qwen3.8 Max was released on 3 Aug 2026.