BenchLeader

GLM 5.3 vs GPT-5.2

Full profiles: GLM 5.3 (Zhipu AI) and GPT-5.2 (OpenAI): every benchmark, price and rank.

Updated · Data from 12 sources

As of 16 Nov 2026, GLM 5.3 and GPT-5.2 are neck and neck on quality (58.8 vs 59.9 on the BenchLeader Index, inside the ±5.0 margin), and GLM 5.3 is 2.2× cheaper at $2.15 vs $4.81 per 1M tokens blended. Choose GLM 5.3 for agents & tools (58 vs 54) and maths (60 vs 57) and fast first answers (57.6 s vs 112.5 s); choose GPT-5.2 for reasoning (60 vs 55), coding (63 vs 60), and knowledge (56 vs 54).

Key facts (as of 16 Nov 2026)

  • GLM 5.3 scores 58.8 on the BenchLeader Index and GPT-5.2 scores 59.9 (BenchLeader, 16 Nov 2026).
  • GLM 5.3 costs $2.15 per 1M tokens blended (list price, 16 Nov 2026).
  • GPT-5.2 costs $4.81 per 1M tokens blended (list price, 16 Nov 2026).
  • Of the 23 benchmarks both models have, GLM 5.3 leads 14 (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 60 category score; LiveCodeBench 85.4% vs 80.5%Medium
Agents & tool useGLM 5.358 vs 54 category scoreMedium
MathsGLM 5.360 vs 57 category score; FrontierMath Tiers 1–3 68.8% vs 67.4%Medium
ReasoningGPT-5.260 vs 55 category score; GPQA Diamond 91.4% vs 90.9%Medium
KnowledgeGPT-5.256 vs 54 category score; TaxEval 75.8% vs 72.4%Medium
Long documentsGPT-5.266 vs 65 category score; AA-LCR v1.1 82.7% vs 79.7%Medium
Chat quality (human preference)–Only GLM 5.3 has an Arena rating (67 category score)Not comparable
Lowest costGLM 5.3$2.15 vs $4.81 per 1M blended; agentic coding session $0.102 vs $0.161–
Fastest first answerGLM 5.357.6 s (GLM 5.3 (max)) vs 112.5 s (GPT-5.2 (xhigh)) 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

  • GLM 5.3 (max) and GPT-5.2 (xhigh) are neck and neck on quality (58.8 vs 59.9, within the ±5.0 margin).
  • GLM 5.3 (max) is stronger in agents & tools, maths.
  • GPT-5.2 (xhigh) is stronger in coding, knowledge, long context, reasoning.
  • GLM 5.3 (max) is 2.2× cheaper ($2.15 vs $4.81 per 1M blended).

At a glance

SpecGLM 5.3GPT-5.2
DeveloperZhipu AIOpenAI
Released18 Aug 202611 Dec 2025
Context window (tokens)1,000,000400,000
Compared atGLM 5.3 (max)GPT-5.2 (xhigh)
Open weightsYesNo
Price in / out / cached input per 1M$1.40 / $4.40 / $0.260$1.75 / $14.00 / $0.175

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.

CategoryGLM 5.3 (max)GPT-5.2 (xhigh)Leader
Reasoning54.760.4GPT-5.2
Coding59.763.0GPT-5.2
Agents & tools57.653.8GLM 5.3
Maths60.056.5GLM 5.3
Knowledge53.556.1GPT-5.2
Instruction following–73.7not comparable
Human preference67.2–not comparable
Long context64.566.0GPT-5.2

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 contextGLM 5.3, Reasoning: 54.7GLM 5.3, Coding: 59.7GLM 5.3, Agents & tools: 57.6GLM 5.3, Maths: 60.0GLM 5.3, Knowledge: 53.5GLM 5.3, Long context: 64.5GPT-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: 66.0

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 23 benchmarks both have, GLM 5.3 leads 14, GPT-5.2 leads 9. Every overall index that rates both puts GLM 5.3 ahead.

Overall indexGLM 5.3 (max)GPT-5.2 (xhigh)Leader
BenchLeader Index58.859.9neck and neck
Arena Text14821437†GLM 5.3

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.

MeasureGLM 5.3 (max)GPT-5.2 (xhigh)
Blended price per 1M tokens$2.15$4.81
Input / output per 1M$1.40 / $4.40$1.75 / $14.00
Cached input per 1M$0.260$0.175
Price per task: chat reply$0.0019$0.0049
Price per task: summarize a 30-page report$0.019$0.029
Price per task: code edit$0.015$0.032
Price per task: agentic coding session$0.102$0.161
Price per task: structured extraction$0.0037$0.0063
Agentic coding session, input mostly cached$0.050$0.090
Cost per benchmark task (AA Index)$5.01–
Output tokens per task (AA Index)119k–
Of which reasoning85k–
Time per task (AA Index)23 min–
Output speed88 tok/s76 tok/s
Time to first answer57.6 s112.5 s
Time to a complete answer, last measured 13 Nov 2026144.1 s154.4 s

Each model is compared at its best-ranked reasoning-effort setting. GLM 5.3 (max) scores 58.8 and takes 58 s to its first answer token; GLM 5.3 (low) scores 47.3 in 7 s. 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. 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

195 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.

MetricPublisherGLM 5.3 (max)GPT-5.2 (xhigh)
BenchLeader IndexBenchLeader58.859.9
Agents & tools scoreBenchLeader57.653.8
Coding scoreBenchLeader59.763.0
Human preference scoreBenchLeader67.2–
Knowledge scoreBenchLeader53.556.1
Long context scoreBenchLeader64.566.0
Maths scoreBenchLeader60.056.5
Reasoning scoreBenchLeader54.760.4
Instruction following scoreBenchLeader–73.7
Blended price $/MBenchLeader$2.15$4.81
Output speedBenchLeader88 tok/s76 tok/s
Time to first answerBenchLeader57.6 s112.5 s
Context windowBenchLeader1,000,000 tokens400,000 tokens
Reasoning
GPQA DiamondEpoch AI Benchmarking Hub90.9%91.4%
Humanity's Last ExamScale AI / CAIS–27.8%†
SimpleBenchSimpleBench66.2%†45.8%†
Arena Hard PromptsArena (LMArena)15071497†
LiveBench 1.0 ReasoningLiveBench85.8%†83.2%†
GPQA Diamond (AA)Artificial Analysis91.7%90.3%
Humanity's Last Exam (AA)Artificial Analysis42.3%37.7%
HLE-Diamond (AA)Artificial Analysis17.9%–
GPQA Diamond (Vals)Vals AI88.1%91.7%
Kagi LLM BenchmarkKagi LLM Benchmark–73.3%†
ARC-AGI-1ARC Prize–86.2%
ARC-AGI-2ARC Prize–52.9%
CritPtArtificial Analysis19.1%11.6%
CritPt v2 hard subset (AA)Artificial Analysis33.5%–
MysteryMechanismVals AI23.0%–
SciPredictScale AI SEAL–20.6%†
HLE-Diamond (Scale)Scale AI SEAL16.4%†–
Arena ExpertArena (LMArena)15221491†
Chess PuzzlesEpoch AI Benchmarking Hub37.0%48.0%
Mystery Game PuzzlesEpoch AI Benchmarking Hub33.0%22.0%†
CL-benchEpoch AI Benchmarking Hub–18.2%†
LMCAEpoch AI Benchmarking Hub55.5%†43.9%
DTBenchEpoch AI Benchmarking Hub87.7%†90.9%
ForecastBenchForecastBench–63.8%†
BTF-3Epoch AI Benchmarking Hub14.9%†–
Show all 195 benchmarks
MetricPublisherGLM 5.3 (max)GPT-5.2 (xhigh)
Coding
SWE-bench Verified (Epoch)Epoch AI Benchmarking Hub–73.8%†
SciCodeSciCode59.0%–
WeirdMLWeirdML75.4%72.2%
FrontierCodeCognition40.1%–
GSO-BenchGSO-Bench–27.4%†
Arena CodingArena (LMArena)15221515†
Arena WebDevArena (LMArena)16231421†
LiveBench 1.0 CodingLiveBench79.0%†76.1%†
LiveCodeBenchVals AI80.5%85.4%
IOIVals AI68.4%–
SWE-bench (Vals)Vals AI95.4%75.8%
SWE-Bench ProScale AI SEAL–29.9%†
SWE-bench Verified (bash only)SWE-bench–69.0%†
SWE-bench Multilingual (bash only)SWE-bench–66.7%†
SWE-bench Verified (any scaffold)SWE-bench–72.8%†
Code MigrationVals AI44.2%–
Code Migration v1.1Vals AI55.1%–
ProgramBenchVals AI1.5%–
Vibe Code Bench 1-100Vals AI20.0%–
Vibe Code Bench v1.1Vals AI78.1%53.5%
Vibe Code Bench 1-100 v1.1Vals AI29.8%–
SWE-Bench Pro (private)Scale AI SEAL–23.8%†
SWE Atlas v1.1Scale AI SEAL58.1%†–
SWE Atlas v1.1: Codebase QnAScale AI SEAL70.2%†–
SWE Atlas v1.1: Test WritingScale AI SEAL55.6%†–
SWE Atlas v1.1: RefactoringScale AI SEAL48.6%†–
SWE-Bench Pro V2Scale AI SEAL95.6%–
SWE-Bench Pro V2 HARDScale AI SEAL84.3%–
MirrorCodeEpoch AI Benchmarking Hub–32.2%
DeepSWE v1.1Epoch AI Benchmarking Hub69.0%–
CursorBenchCursor42.6%–
ALE-BenchEpoch AI Benchmarking Hub1317.4†1249.8†
AlgoTuneAlgoTune–2.0†
FrontierSWEEpoch AI Benchmarking Hub30.2%–
WeirdML v3WeirdML–5.0%
Terminal-Bench 4.0 (AA)Artificial Analysis41.9%–
Terminal-Bench 2.1 (AA)Artificial Analysis83.9%–
Startrise FrontendStartrise73.2%†–
GameBench 2 OverallDuelLab GameBench 264.6†–
GameBench 2 by reasoning settingDuelLab GameBench 272.3–
Design Arena WebsiteDesign Arena1279†1186
FrontierCode 1.1 (AA)Artificial Analysis40.1%–
FrontierCode 2.0 (AA)Artificial Analysis21.6%–
LiveCodeBench (AA)Artificial Analysis–88.9%
Agents & tools
Terminal-BenchTerminal-Bench41.2%64.9%†
GDPvalOpenAI–49.7%†
Remote Labor IndexScale AI / CAIS–2.1%†
APEX-AgentsMercor56.6%†–
Arena AgentArena (LMArena)1–
LiveBench Agentic CodingLiveBench60.9%†50.3%†
Terminal-Bench HardArtificial Analysis–47.0%
τ²-Bench Telecom (AA)Artificial Analysis–84.8%
Terminal-Bench 2.1 (Vals)Vals AI71.5%–
MCP AtlasScale AI SEAL84.2%†67.6%
HiL-BenchScale AI SEAL58.3%†–
BFCL OverallBerkeley Function Calling Leaderboard–55.9%†
τ²-benchτ²-bench–84.8%†
τ³-bench Bankingτ²-bench–32.2%†
GDPval-AA v2.1, normalised scoreArtificial Analysis57.5%–
GDPval-AA v2.1Artificial Analysis1655–
τ³-Banking (AA)Artificial Analysis50.3%11.1%†
ITBench SRE (AA)Artificial Analysis46.1%–
Analyst Agent (AA)Artificial Analysis38.8%–
CyberBenchVals AI72.1%–
Finance Agent v2Vals AI55.8%–
Finance Agent v2.1Vals AI65.0%–
Harvey's Legal Agent Benchmark v1.1Vals AI14.6%–
Harvey's Legal Agent Benchmark v1Vals AI8.3%–
Legal Research BenchVals AI49.0%–
Poker AgentVals AI–1131.8%†
SkillsBenchVals AI47.5%–
SREBenchVals AI9.2%–
Tax Agent BenchVals AI43.3%–
Terminal-Bench 2.0 (Vals)Vals AI–51.7%†
Terminal-Bench 4.0 (Vals)Vals AI38.9%–
Terminal-Bench ScienceVals AI5.7%–
Arena SearchArena (LMArena)–1161†
Arena Agent Task CompletionArena (LMArena)3.1–
DeepResearch BenchDeepResearch Bench–41.1%†
PostTrainBenchEpoch AI Benchmarking Hub34.5%–
METR Time HorizonsMETR–75.3%†
Vending-Bench 2Andon Labs8163.6†3591.3†
AutomationBenchArtificial Analysis62.2%–
GDP.pdf (AA)Artificial Analysis11.2%–
Harvey LAB-AA v1.1Harvey0.3%–
EnterpriseOps-GymArtificial Analysis36.4%–
AA-Briefcase v1.1Artificial Analysis1501–
GDPxGDPx by SemiAnalysis2.7%–
Design Arena AgenticDesign Arena1239†1065†
Terminal-Bench-Science 0.1 (AA)Artificial Analysis9.5%–
CWE-Bench (AA)Artificial Analysis55.8%–
DeepSec-Bench (AA)Artificial Analysis24.3%–
CyberGym end-to-end (AA)Artificial Analysis29.0%–
FACTS SearchGoogle87.7%†–
Maths
FrontierMath Tiers 1–3Epoch AI Benchmarking Hub68.8%67.4%
FrontierMath Tier 4Epoch AI Benchmarking Hub29.3%31.7%
OTIS Mock AIMEEpoch AI Benchmarking Hub91.1%96.1%
ProofBenchVals AI49.0%15.0%
LiveBench MathematicsLiveBench87.9%†93.2%†
AIME (Vals)Vals AI–96.9%
MGSMVals AI–94.0%
AIME 2026MathArena–98.3%†
HMMT February 2026MathArena–97.0%†
MathArena ApexMathArena–13.5%†
Arena MathsArena (LMArena)15011453†
AIME 2025 (AA)Artificial Analysis–99.0%
Knowledge
SimpleQA VerifiedEpoch AI Benchmarking Hub41.0%37.1%
LiveBench 1.0 Data AnalysisLiveBench70.2%†78.2%†
AA-Omniscience v1.1Artificial Analysis6.0–
MMLU-ProVals AI86.8%86.2%
LegalBenchVals AI84.8%82.8%
CorpFinVals AI–65.9%
TaxEvalVals AI72.4%75.8%
MedQAVals AI–94.1%
MultiNRCScale AI SEAL–42.2%†
CaseLaw v2Vals AI–66.0%
Excel Modeling BenchmarkVals AI56.3%–
Excel Modeling Benchmark v1.1Vals AI69.1%–
MedCodeVals AI42.9%49.8%
MedScribeVals AI88.8%84.4%
Public Benefits Bench v1.1Vals AI68.5%–
Access to Justice BenchmarkVals AI73.7%–
Financial Coaching BenchVals AI74.1%–
Immigration BenchVals AI66.7%–
DrugDiscoveryBenchScale AI SEAL–29.3%
EBR-benchEpoch AI Benchmarking Hub–23.0%
AA-Omniscience v1.1: accuracyArtificial Analysis32.3%–
AA-Omniscience v1.1: non-hallucinationArtificial Analysis61.1%–
Global-MMLU-Lite (AA)Artificial Analysis–89.8%†
HealthBench Professional (AA)OpenAI52.8%–
FACTS ParametricGoogle40.3%†–
Instruction following
LiveBench LanguageLiveBench79.9%†79.8%†
LiveBench Instruction FollowingLiveBench69.3%†61.8%†
IFBenchArtificial Analysis–75.4%
TutorBenchScale AI SEAL–53.5%†
Arena Instruction FollowingArena (LMArena)14811457†
Human preference
Arena TextArena (LMArena)14821476†
EQ-Bench Creative Writing v3EQ-Bench2076†1703†
Arena Creative WritingArena (LMArena)14561432†
Arena Multi-turnArena (LMArena)14821495†
Arena Longer QueriesArena (LMArena)14881472†
Multimodal
Arena VisionArena (LMArena)–1268†
MMMU-ProArtificial Analysis–74.6%†
VISTAScale AI SEAL–46.6%†
MMMU-Pro (Vals)Vals AI–86.7%
MortgageTaxVals AI–67.1%
VPCTEpoch AI Benchmarking Hub–84.0%
Furniture AssemblyEpoch AI Benchmarking Hub–38.3%
Code2Video-AAArtificial Analysis1034–
AA-3DArtificial Analysis794–
Long context
AA-LCR v1.1Artificial Analysis79.7%82.7%
Arena DocumentArena (LMArena)–1425†
MLCRArtificial Analysis17.2%–
FACTS GroundingGoogle55.2%†–
Composite
Epoch Capabilities IndexEpoch AI Benchmarking Hub157.1†154.5†
LiveBench 1.0LiveBench76.1%†74.6%†
AA Intelligence Index v5.0.1Artificial Analysis31.6–
AA Intelligence Index v4.3.2Artificial Analysis44.830.4
AA Progress Index v1Artificial Analysis322.7295.5
Vals IndexVals AI59.9–
Vals Finance IndexVals AI61.4%–
Vals RSI IndexVals AI22.9%15.8%
Vals Router IndexVals AI59.7%†–
Vals Legal IndexVals AI74.6%–
AA Cyber IndexArtificial Analysis36.4–
Safety & honesty
FORTRESSScale AI SEAL28.2%17.5%†
MASKScale AI SEAL–86.7%†
PropensityBenchScale AI SEAL–34.4%†
DistressBenchScale AI SEAL72.5%†–
Metadata
AA Openness IndexArtificial Analysis33.3–

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

  • GPT-5.2 (xhigh): 1 new result (AIME 2025 (AA)).
  • GLM 5.3 (max): 1 new result (PostTrainBench).
  • GPT-5.2 (xhigh): 1 new result (MirrorCode).
  • GLM 5.3 (max): 4 score revisions.
  • GLM 5.3 (max): 1 score revision; cached input price moved from $0.30/M to $0.26/M (13% down).
  • GLM 5.3 (max): cached input price moved from $0.26/M to $0.30/M (15% up).
  • GLM 5.3 (max): cached input price moved from $0.30/M to $0.26/M (13% down).
  • GLM 5.3 (max): cached input price moved from $0.26/M to $0.30/M (15% up).
  • GLM 5.3 (max): BenchLeader Index 61.3 → 58.9.
  • GLM 5.3 (max): 5 new results (AA Intelligence Index v5.0.1, AA Progress Index v1, HLE-Diamond (AA)…); 3 score revisions; cost per task moved from $2.01/M to $5.01/M (150% up); cached input price moved from $0.30/M to $0.26/M (13% down).

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

GLM 5.3 vs GPT-5.2: questions

Is GLM 5.3 better than GPT-5.2?
GLM 5.3 (max) and GPT-5.2 (xhigh) are neck and neck on quality (58.8 vs 59.9, within the ±5.0 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.
GLM 5.3 vs GPT-5.2: which is cheaper, and will that change?
GLM 5.3 is cheaper today: $2.15 against $4.81 per million tokens, blended at three input tokens per output token. No price change has been announced for either.
Is GLM 5.3 or GPT-5.2 better for coding?
GPT-5.2 scores higher in coding (63 vs 60 on the category index, where 50 is average). On LiveCodeBench, the top shared coding benchmark, it is 80.5% for GLM 5.3 against 85.4% for GPT-5.2.
Which is faster, GLM 5.3 or GPT-5.2?
GLM 5.3 streams faster (88 vs 76 output tokens per second, Artificial Analysis 7-day median).
Is GLM 5.3 or GPT-5.2 better for agents and tool use?
GLM 5.3 scores higher in agentic tasks (58 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). GLM 5.3 accepts more context: 1,000,000 against 400,000 tokens for GPT-5.2.
Which has lower latency?
GLM 5.3 answers first: 57.6 s against 112.5 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 GLM 5.3 vs GPT-5.2?
Across the 23 benchmarks both have, GLM 5.3 leads 14, GPT-5.2 leads 9. Every overall index that rates both puts GLM 5.3 ahead.
When was GLM 5.3 released?
GLM 5.3 was released on 18 Aug 2026.