BenchLeader
Menu
Best LLMsChoose a modelCoding plansLeaderboardsModelsBenchmarksProvidersReleases8

GPT-5.6 Sol vs Qwen3.7

Full profiles: GPT-5.6 Sol (OpenAI) and Qwen3.7 (Alibaba): every benchmark, price and rank.

Updated · Data from 12 sources

As of 29 Oct 2026, GPT-5.6 Sol is ahead of Qwen3.7 on quality (65.4 vs 60.3 on the BenchLeader Index), but Qwen3.7 is 2.1× cheaper at $3.75 vs $8.00 per 1M tokens blended. Choose GPT-5.6 Sol for agents & tools (67 vs 49), maths (71 vs 58), and coding (66 vs 59); choose Qwen3.7 for fast first answers (14.1 s vs 86.2 s) and instruction following (78 vs 71) and reasoning (66 vs 65), where it also streams ~2× faster.

Key facts (as of 29 Oct 2026)

  • GPT-5.6 Sol scores 65.4 on the BenchLeader Index and Qwen3.7 scores 60.3 (BenchLeader, 29 Oct 2026).
  • GPT-5.6 Sol costs $8.00 per 1M tokens blended (list price, 29 Oct 2026).
  • Qwen3.7 costs $3.75 per 1M tokens blended (list price, 29 Oct 2026).
  • Qwen3.7 streams ~2× faster than GPT-5.6 Sol (Artificial Analysis, 7-day median, 29 Oct 2026).
  • Of the 56 benchmarks both models have, GPT-5.6 Sol leads 49 (BenchLeader, 29 Oct 2026).
GPT-5.6 Sol benchmarks max OpenAI Qwen3.7 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-5.6 Sol66 vs 59 category score; SciCode 57.1% vs 49.5%Medium
Agents & tool useGPT-5.6 Sol67 vs 49 category score; τ³-Banking (AA) 44.3% vs 11.8%High
MathsGPT-5.6 Sol71 vs 58 category score; FrontierMath Tiers 1–3 89.1% vs 64.6%High
ReasoningQwen3.766 vs 65 category score; GPQA Diamond 90.9% vs 93.5%Medium
KnowledgeGPT-5.6 Sol64 vs 61 category score; SimpleQA Verified 69.7% vs 55.8%Medium
Long documentsGPT-5.6 Sol67 vs 64 category score; AA-LCR v1.1 84.0% vs 79.0%Medium
Following instructionsQwen3.778 vs 71 category score; IFBench 80.5% vs 72.7%Medium
Chat quality (human preference)–Only Qwen3.7 has an Arena rating (56 category score)Not comparable
Lowest costQwen3.7$3.75 vs $8.00 per 1M blended; agentic coding session $0.180 vs $0.320–
Fastest first answerQwen3.714.1 s (Qwen3.7 (max)) vs 86.2 s (GPT-5.6 Sol (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-5.6 Sol (max) leads on quality: 65.4 vs 60.3.
  • GPT-5.6 Sol (max) is stronger in agents & tools, coding, knowledge, long context, maths, multimodal.
  • Qwen3.7 (max) is stronger in instruction following, reasoning, human preference.
  • Qwen3.7 (max) is 2.1× cheaper ($3.75 vs $8.00 per 1M blended).
  • Qwen3.7 (max) streams 2.4× faster (201 vs 85 tokens per second).

At a glance

SpecGPT-5.6 SolQwen3.7
DeveloperOpenAIAlibaba
Released9 Jul 202619 May 2026
Context window (tokens)1,050,0001,000,000
Compared atGPT-5.6 Sol (max)Qwen3.7 (max)
Open weightsNoNo
Price in / out / cached input per 1M$4.00 / $20.00 / $0.400$2.50 / $7.50 / $0.330

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-5.6 Sol (max)Qwen3.7 (max)Leader
Reasoning65.166.2Qwen3.7
Coding65.859.2GPT-5.6 Sol
Agents & tools66.649.2GPT-5.6 Sol
Maths70.858.2GPT-5.6 Sol
Knowledge63.961.2GPT-5.6 Sol
Instruction following71.178.0Qwen3.7
Human preference–55.7not comparable
Multimodal66.2–not comparable
Long context66.964.3GPT-5.6 Sol

Where sources disagree

Across the 56 benchmarks both have, GPT-5.6 Sol leads 49, Qwen3.7 leads 7. Every overall index that rates both puts GPT-5.6 Sol ahead.

Overall indexGPT-5.6 Sol (max)Qwen3.7 (max)Leader
BenchLeader Index65.460.3GPT-5.6 Sol
Arena Text1485†1474GPT-5.6 Sol

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-5.6 Sol (max)Qwen3.7 (max)
Blended price per 1M tokens$8.00$3.75
Input / output per 1M$4.00 / $20.00$2.50 / $7.50
Cached input per 1M$0.400$0.330
Price per task: chat reply$0.0076$0.0033
Price per task: summarize a 30-page report$0.060$0.035
Price per task: code edit$0.054$0.026
Price per task: agentic coding session$0.320$0.180
Price per task: structured extraction$0.012$0.0065
Output speed85 tok/s201 tok/s
Time to first answer86.2 s14.1 s
Time to a complete answer92.1 s16.6 s

Each model is compared at its best-ranked reasoning-effort setting. GPT-5.6 Sol (max) scores 65.4 and takes 86 s to its first answer token; GPT-5.6 Sol (xhigh) scores 64.4 in 22 s, GPT-5.6 Sol (high) scores 63.6 in 11 s. Qwen3.7 (max) scores 60.3 and takes 14 s to its first answer token. 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

130 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-5.6 Sol (max)Qwen3.7 (max)
BenchLeader IndexBenchLeader65.460.3
Agents & tools scoreBenchLeader66.649.2
Coding scoreBenchLeader65.859.2
Instruction following scoreBenchLeader71.178.0
Knowledge scoreBenchLeader63.961.2
Long context scoreBenchLeader66.964.3
Maths scoreBenchLeader70.858.2
Multimodal scoreBenchLeader66.2–
Reasoning scoreBenchLeader65.166.2
Human preference scoreBenchLeader–55.7
Blended price $/MBenchLeader$8.00$3.75
Output speedBenchLeader85 tok/s201 tok/s
Time to first answerBenchLeader86.2 s14.1 s
Context windowBenchLeader1,050,000 tokens1,000,000 tokens
Reasoning
GPQA DiamondEpoch AI Benchmarking Hub93.5%90.9%
SimpleBenchSimpleBench64.8%†70.4%
Arena Hard PromptsArena (LMArena)1512†1495
LiveBench 1.0 ReasoningLiveBench91.7%83.3%
GPQA Diamond (AA)Artificial Analysis94.1%92.3%
Humanity's Last Exam (AA)Artificial Analysis49.5%40.5%
GPQA Diamond (Vals)Vals AI95.2%90.2%
EnigmaEvalScale AI SEAL37.1%†–
Kagi LLM BenchmarkKagi LLM Benchmark67.0%†–
ARC-AGI-1ARC Prize96.5%–
ARC-AGI-2ARC Prize92.5%–
ARC-AGI-3ARC Prize7.8%–
CritPtArtificial Analysis32.3%13.4%
CritPt v2 hard subset (AA)Artificial Analysis56.0%–
MysteryMechanismVals AI33.3%–
Chess PuzzlesEpoch AI Benchmarking Hub59.0%26.0%
Mystery Game PuzzlesEpoch AI Benchmarking Hub58.0%32.0%
LMCAEpoch AI Benchmarking Hub58.4%44.0%
DTBenchEpoch AI Benchmarking Hub95.5%92.3%
ForecastBenchForecastBench65.5%–
BTF-3Epoch AI Benchmarking Hub13.7%–
Coding
SWE-bench Verified (Epoch)Epoch AI Benchmarking Hub–77.3%
SciCodeSciCode57.1%49.5%
WeirdMLWeirdML87.0%–
FrontierCodeCognition47.5%†–
Show all 130 benchmarks
MetricPublisherGPT-5.6 Sol (max)Qwen3.7 (max)
Coding
GSO-BenchGSO-Bench76.5%†–
Arena CodingArena (LMArena)1534†1524
Arena WebDevArena (LMArena)1616†1515
LiveBench 1.0 CodingLiveBench83.9%74.2%
LiveCodeBenchVals AI82.6%87.1%
IOIVals AI91.2%–
SWE-bench (Vals)Vals AI96.2%68.8%
Code MigrationVals AI52.9%13.1%
ProgramBenchVals AI1.5%–
Vibe Code Bench 1-100Vals AI20.0%–
Vibe Code Bench v1.1Vals AI80.5%47.7%
SWE Atlas: Codebase QnAScale AI SEAL46.0%†–
SWE Atlas: Test WritingScale AI SEAL45.9%†–
MirrorCodeEpoch AI Benchmarking Hub20.0%†–
Surface Evolver BenchEpoch AI Benchmarking Hub93.1%†–
DeepSWE v1.1Epoch AI Benchmarking Hub72.7%–
CursorBenchCursor41.7%–
ALE-BenchEpoch AI Benchmarking Hub2176.91189.4
FrontierSWEEpoch AI Benchmarking Hub32.2%–
Terminal-Bench 4.0 (AA)Artificial Analysis39.9%1.5%
Terminal-Bench 2.1 (AA)Artificial Analysis88.0%74.5%
Startrise FrontendStartrise66.2%†63.6%
Agents & tools
Terminal-BenchTerminal-Bench37.3%–
OSWorld-Verified 2.0OSWorld27.3%–
Arena AgentArena (LMArena)6.1†-6.4
LiveBench Agentic CodingLiveBench56.2%43.6%
Terminal-Bench HardArtificial Analysis65.9%50.8%
τ²-Bench Telecom (AA)Artificial Analysis85.1%94.7%
Terminal-Bench 2.1 (Vals)Vals AI85.8%61.0%
HiL-BenchScale AI SEAL32.3%†–
τ³-bench Bankingτ²-bench46.9%†–
τ³-Banking (AA)Artificial Analysis44.3%11.8%
ITBench SRE (AA)Artificial Analysis56.2%42.5%
Analyst Agent (AA)Artificial Analysis47.5%18.8%
CUA-benchVals AI8.3%–
CyberBenchVals AI76.3%–
Finance Agent v2Vals AI53.8%47.8%
Harvey's Legal Agent Benchmark v1Vals AI2.5%1.7%
Legal Research BenchVals AI48.1%25.5%
SkillsBenchVals AI54.1%–
SREBenchVals AI30.5%–
Tax Agent BenchVals AI31.1%15.8%
Terminal-Bench 2.0 (Vals)Vals AI–59.2%
Terminal-Bench 4.0 (Vals)Vals AI37.9%–
Terminal-Bench ScienceVals AI20.0%–
Time Horizon Index: KSPVals AI23.8%–
Web Search IndexVals AI43.6%–
Arena SearchArena (LMArena)1258†–
BALROGBALROG60.0%–
PostTrainBenchEpoch AI Benchmarking Hub36.2%–
GBAEvalEpoch AI Benchmarking Hub52.6%†0.4%
Vending-Bench 2Andon Labs9619.4†–
AutomationBenchArtificial Analysis60.1%23.0%
GDP.pdf (AA)Artificial Analysis27.2%8.6%
EnterpriseOps-GymArtificial Analysis42.9%45.0%
AA-Briefcase v1.1Artificial Analysis1478917
Terminal-Bench-Science 0.1 (AA)Artificial Analysis22.4%–
FACTS SearchGoogle88.5%†–
Maths
FrontierMath Tiers 1–3Epoch AI Benchmarking Hub89.1%64.6%
FrontierMath Tier 4Epoch AI Benchmarking Hub82.9%34.1%
OTIS Mock AIMEEpoch AI Benchmarking Hub100.0%95.6%
ProofBenchVals AI83.0%26.0%
LiveBench MathematicsLiveBench96.2%85.3%
Arena MathsArena (LMArena)1496†1484
Knowledge
SimpleQA VerifiedEpoch AI Benchmarking Hub69.7%55.8%
LiveBench 1.0 Data AnalysisLiveBench79.8%71.8%
MMLU-ProVals AI89.1%89.3%
LegalBenchVals AI87.0%84.9%
CorpFinVals AI64.4%63.7%
TaxEvalVals AI74.8%75.3%
PRBench FinanceScale AI SEAL50.5%–
PRBench LegalScale AI SEAL50.5%–
BioMysteryBenchVals AI71.1%–
Excel Modeling BenchmarkVals AI72.3%57.0%
MedCodeVals AI44.0%38.8%
MedScribeVals AI85.2%79.4%
Public Benefits Bench v1.1Vals AI66.5%–
DrugDiscoveryBenchScale AI SEAL56.5%29.3%†
EBR-benchEpoch AI Benchmarking Hub44.8%9.5%
GDP.pdfEpoch AI Benchmarking Hub30.7%–
Instruction following
LiveBench LanguageLiveBench87.7%79.7%
LiveBench Instruction FollowingLiveBench71.8%74.0%
IFBenchArtificial Analysis72.7%80.5%
Arena Instruction FollowingArena (LMArena)1491†1466
Human preference
Arena TextArena (LMArena)1485†1474
EQ-Bench 4EQ-Bench1236†1100
Arena Creative WritingArena (LMArena)1470†1438
Arena Multi-turnArena (LMArena)1490†1482
Arena Longer QueriesArena (LMArena)1499†1491
Multimodal
Arena VisionArena (LMArena)1281†–
MMMU-ProArtificial Analysis83.4%–
MMMU-Pro (Vals)Vals AI88.8%–
MortgageTaxVals AI67.3%–
Blueprint-Bench 2Epoch AI Benchmarking Hub33.6%†–
Long context
AA-LCR v1.1Artificial Analysis84.0%79.0%
Arena DocumentArena (LMArena)1489†–
MLCRArtificial Analysis30.0%14.4%
Composite
Epoch Capabilities IndexEpoch AI Benchmarking Hub162.8†154.0
LiveBench 1.0LiveBench81.0%73.1%
AA Intelligence Index v4.3.2Artificial Analysis47.029.5
AA Progress Index v1Artificial Analysis331.5293.8
Vals RSI IndexVals AI23.9%–
Vals Science IndexVals AI41.5%–
Vals Multimodal IndexVals AI72.6%–
Safety & honesty
FORTRESSScale AI SEAL20.1%†–

Data as of 2026-10-29. Best configuration of each model; every score links to its source on the model pages.

What changed

  • Qwen3.7 (max): cached input price moved from $0.50/M to $0.33/M (34% down).
  • Qwen3.7 (max): BenchLeader Index 60.8 → 60.3.
  • GPT-5.6 Sol (max): 2 new results (AA Progress Index v1, CritPt v2 hard subset (AA)).
  • Qwen3.7 (max): 1 new result (AA Progress Index v1).
  • Qwen3.7 (max): cached input price moved from $0.33/M to $0.50/M (52% up).
  • Qwen3.7 (max): cost per task moved from $1.15/M to $0.93/M (19% down).
  • GPT-5.6 Sol (max): 1 new result (AA-Briefcase v1.1).
  • Qwen3.7 (max): 1 new result (AA-Briefcase v1.1); 1 score revision; cached input price moved from $0.50/M to $0.33/M (34% down).
  • Qwen3.7 (max): cached input price moved from $0.33/M to $0.50/M (52% up).
  • GPT-5.6 Sol (max): 3 new results (Harvey's Legal Agent Benchmark v1, Vals RSI Index, Vals Science Index).

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

GPT-5.6 Sol vs Qwen3.7: questions

Is GPT-5.6 Sol better than Qwen3.7?
GPT-5.6 Sol (max) leads on quality: 65.4 vs 60.3. The BenchLeader Index combines every independent quality benchmark; GPT-5.6 Sol (max) is ahead overall as of 29 Oct 2026, but check the category scores for your use.
GPT-5.6 Sol vs Qwen3.7: which is cheaper, and will that change?
Qwen3.7 is cheaper today: $3.75 against $8.00 per million tokens, blended at three input tokens per output token. No price change has been announced for either.
Is GPT-5.6 Sol or Qwen3.7 better for coding?
GPT-5.6 Sol scores higher in coding (66 vs 59 on the category index, where 50 is average). On LiveCodeBench, the top shared coding benchmark, it is 82.6% for GPT-5.6 Sol against 87.1% for Qwen3.7.
Which is faster, GPT-5.6 Sol or Qwen3.7?
Qwen3.7 streams ~2× faster (201 vs 85 output tokens per second, Artificial Analysis 7-day median).
Is GPT-5.6 Sol or Qwen3.7 better for agents and tool use?
GPT-5.6 Sol scores higher in agentic tasks (67 vs 49 on the category index, where 50 is average). On τ³-Banking (AA), the top shared agents & tools benchmark, it is 44.3% for GPT-5.6 Sol against 11.8% for Qwen3.7.
More questions (4)
Which is better for long documents?
GPT-5.6 Sol scores higher on long-context benchmarks (67 vs 64). Both take about 1.1M tokens: GPT-5.6 Sol 1,050,000, Qwen3.7 1,000,000.
Which has lower latency?
Qwen3.7 answers first: 14.1 s against 86.2 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-5.6 Sol vs Qwen3.7?
Across the 56 benchmarks both have, GPT-5.6 Sol leads 49, Qwen3.7 leads 7. Every overall index that rates both puts GPT-5.6 Sol ahead.
When was GPT-5.6 Sol released?
GPT-5.6 Sol was released on 9 Jul 2026.