BenchLeader
Menu
Best LLMsChoose a modelLeaderboardsModelsBenchmarksProvidersReleases9

Claude Opus 5 vs GLM 5.3

Full profiles: Claude Opus 5 (Anthropic) and GLM 5.3 (Zhipu AI): every benchmark, price and rank.

Updated · Data from 11 sources

As of 26 Oct 2026, Claude Opus 5 (max) leads on quality (68.8 vs 63.5 on the BenchLeader Index), but GLM 5.3 (max) costs 4.7× less, so it is the better buy unless you need the extra quality. Claude Opus 5 leads in every category we measure both in.

Claude Opus 5 benchmarks max Anthropic GLM 5.3 benchmarks max Zhipu AIOpen 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 570 vs 62 category score; Arena WebDev 1691 vs 1623High
Agents & tool useClaude Opus 564 vs 58 category score; Terminal-Bench 51.8% vs 41.8%Medium
MathsClaude Opus 571 vs 60 category score; FrontierMath Tiers 1–3 85.6% vs 68.8%High
ReasoningClaude Opus 574 vs 69 category score; GPQA Diamond 93.9% vs 90.9%Medium
KnowledgeClaude Opus 572 vs 59 category score; SimpleQA Verified 59.9% vs 41.0%High
Long documentsEither65 vs 65 category score; AA-LCR 79.3% vs 79.7%Either
Chat quality (human preference)Claude Opus 569 vs 67 on Arena votesMedium
Lowest costGLM 5.3$2.15 vs $10.00 per 1M blended; agentic coding session $0.102 vs $0.400–
Fastest first answerGLM 5.322.7 s (GLM 5.3 (max)) vs 68.7 s (Claude Opus 5 (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

  • Claude Opus 5 (max) leads on quality: 68.8 vs 63.5.
  • Claude Opus 5 (max) is stronger in agents & tools, coding, composite, human preference, knowledge, maths, multimodal, reasoning.
  • GLM 5.3 (max) is stronger in long context.
  • GLM 5.3 (max) is 4.7× cheaper ($2.15 vs $10.00 per 1M blended).
  • GLM 5.3 (max) streams 1.9× faster (98 vs 51 tokens per second).

At a glance

SpecClaude Opus 5GLM 5.3
DeveloperAnthropicZhipu AI
Released24 Jul 202618 Aug 2026
Context window (tokens)1,000,0001,000,000
Compared atClaude Opus 5 (max)GLM 5.3 (max)
Open weightsNoYes
Price in / out / cached input per 1M$5.00 / $25.00 / $0.500$1.40 / $4.40 / $0.299

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 5 (max)GLM 5.3 (max)Leader
Reasoning74.168.5Claude Opus 5
Coding70.261.9Claude Opus 5
Agents & tools64.357.9Claude Opus 5
Maths70.660.4Claude Opus 5
Knowledge71.658.5Claude Opus 5
Human preference68.567.3Claude Opus 5
Multimodal67.9–not comparable
Long context64.764.9level
Composite86.379.6Claude Opus 5

Where sources disagree

Across the 43 benchmarks both have, Claude Opus 5 leads 37, GLM 5.3 leads 6. Every overall index that rates both puts Claude Opus 5 ahead.

Overall indexClaude Opus 5 (max)GLM 5.3 (max)Leader
BenchLeader Index68.863.5Claude Opus 5
AA Intelligence Index v4.3.250.844.8Claude Opus 5
Arena Text14911481Claude Opus 5

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 5 (max)GLM 5.3 (max)
Blended price per 1M tokens$10.00$2.15
Input / output per 1M$5.00 / $25.00$1.40 / $4.40
Cached input per 1M$0.500$0.299
Price per task: chat reply$0.0095$0.0019
Price per task: summarize a 30-page report$0.075$0.019
Price per task: code edit$0.068$0.015
Price per task: agentic coding session$0.400$0.102
Price per task: structured extraction$0.015$0.0037
Cost to run a full benchmark suite$5.86$2.01
Output speed51 tok/s98 tok/s
Time to first answer68.7 s22.7 s
Time to a complete answer78.4 s27.8 s

Per token is not per task: on Artificial Analysis' full suite Claude Opus 5 (max) cost $5.86 and GLM 5.3 (max) $2.01, because GLM 5.3 (max) used about 1.6× as many tokens for the same work, which a per-token price hides.

Each model is compared at its best-ranked reasoning-effort setting. Claude Opus 5 (max) scores 68.8 and takes 69 s to its first answer token; Claude Opus 5 (high) scores 67.9 in 17 s, Claude Opus 5 (xhigh) scores 67.4 in 31 s. GLM 5.3 (max) scores 63.5 and takes 23 s to its first answer token; GLM 5.3 (low) scores 55.6 in 25 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

137 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 5 (max)GLM 5.3 (max)
BenchLeader IndexBenchLeader68.863.5
Agents & tools scoreBenchLeader64.357.9
Coding scoreBenchLeader70.261.9
Composite scoreBenchLeader86.379.6
Human preference scoreBenchLeader68.567.3
Knowledge scoreBenchLeader71.658.5
Long context scoreBenchLeader64.764.9
Maths scoreBenchLeader70.660.4
Multimodal scoreBenchLeader67.9–
Reasoning scoreBenchLeader74.168.5
Blended price $/MBenchLeader$10.00$2.15
Output speedBenchLeader51 tok/s98 tok/s
Time to first answerBenchLeader68.7 s22.7 s
Context windowBenchLeader1,000,000 tokens1,000,000 tokens
Reasoning
GPQA DiamondEpoch AI Benchmarking Hub93.9%90.9%
SimpleBenchSimpleBench80.6%†–
Arena Hard PromptsArena (LMArena)15151505
LiveBench 1.0 ReasoningLiveBench91.2%85.8%†
GPQA Diamond (AA)Artificial Analysis93.2%91.7%
Humanity's Last Exam (AA)Artificial Analysis54.9%42.3%
GPQA Diamond (Vals)Vals AI93.4%†88.1%
ARC-AGI-1ARC Prize97.5%–
ARC-AGI-2ARC Prize90.4%–
ARC-AGI-3ARC Prize30.2%†–
CritPtArtificial Analysis29.1%19.1%
MysteryMechanismVals AI37.4%†23.0%
Chess PuzzlesEpoch AI Benchmarking Hub43.0%37.0%
Mystery Game PuzzlesEpoch AI Benchmarking Hub59.0%33.0%
LMCAEpoch AI Benchmarking Hub63.3%55.5%†
DTBenchEpoch AI Benchmarking Hub97.6%87.7%†
ForecastBenchForecastBench66.5%–
BTF-3Epoch AI Benchmarking Hub12.0%†14.9%†
Coding
SciCodeSciCode56.4%59.0%
WeirdMLWeirdML91.8%75.4%
FrontierCodeCognition53.4%40.1%
Arena CodingArena (LMArena)15281521
Arena WebDevArena (LMArena)16911623
LiveBench 1.0 CodingLiveBench81.5%79.0%†
LiveCodeBenchVals AI89.0%†80.5%
Show all 137 benchmarks
MetricPublisherClaude Opus 5 (max)GLM 5.3 (max)
Coding
IOIVals AI84.3%†68.4%
SWE-bench (Vals)Vals AI97.0%†95.4%
Code MigrationVals AI57.5%†44.2%
Code Migration v1.1Vals AI–55.1%
ProgramBenchVals AI3.0%†1.5%
Vibe Code Bench 1-100Vals AI28.5%†20.0%
Vibe Code Bench v1.1Vals AI88.4%†78.1%
SWE Atlas: Codebase QnAScale AI SEAL63.2%†–
SWE Atlas: Test WritingScale AI SEAL62.2%†–
DeepSWE v1.1Epoch AI Benchmarking Hub73.7%69.0%
CursorBenchCursor46.6%42.6%
ALE-BenchEpoch AI Benchmarking Hub2164.6†1317.4†
FrontierSWEEpoch AI Benchmarking Hub52.0%30.2%
Terminal-Bench 4.0 (AA)Artificial Analysis49.0%41.9%
Terminal-Bench 2.1 (AA)Artificial Analysis89.1%83.9%
Startrise FrontendStartrise82.3%†73.2%†
FrontierCode 1.1 (AA)Artificial Analysis–40.1%
Agents & tools
Terminal-BenchTerminal-Bench51.8%41.8%
OSWorld-Verified 2.0OSWorld31.4%–
APEX-AgentsMercor65.8%56.6%†
Arena AgentArena (LMArena)8.21.2
LiveBench Agentic CodingLiveBench65.2%60.9%†
Terminal-Bench 2.1 (Vals)Vals AI84.6%†71.5%
MCP AtlasScale AI SEAL85.8%†84.2%†
HiL-BenchScale AI SEAL57.0%†58.3%†
τ³-bench Bankingτ²-bench48.7%–
GDPval-AA v2.1Artificial Analysis–57.5%
τ³-Banking (AA)Artificial Analysis42.1%50.3%
ITBench SRE (AA)Artificial Analysis–46.1%
Analyst Agent (AA)Artificial Analysis53.8%38.8%
CUA-benchVals AI9.0%–
CyberBenchVals AI65.4%†72.1%
Finance Agent v2Vals AI58.6%†55.8%
Finance Agent v2.1Vals AI–65.0%
Harvey's Legal Agent Benchmark v1.1Vals AI–14.6%
Harvey's Legal Agent Benchmark v1Vals AI6.7%†8.3%
Legal Research BenchVals AI55.3%†49.0%
SkillsBenchVals AI60.4%†47.5%
SREBenchVals AI12.2%†9.2%
Tax Agent BenchVals AI45.5%†39.7%
Terminal-Bench 4.0 (Vals)Vals AI53.5%†38.9%
Terminal-Bench ScienceVals AI27.1%†5.7%
Time Horizon Index: KSPVals AI18.8%†–
Arena SearchArena (LMArena)1214†–
BALROGBALROG63.4%–
PostTrainBenchEpoch AI Benchmarking Hub35.0%†–
GBAEvalEpoch AI Benchmarking Hub79.6%†–
Vending-Bench 2Andon Labs11181.9†8163.6†
AutomationBenchArtificial Analysis56.6%62.2%
GDP.pdf (AA)Artificial Analysis21.6%11.2%
Harvey LAB-AA v1.1Harvey–0.3%
EnterpriseOps-GymArtificial Analysis47.5%36.4%
AA-Briefcase v1.1Artificial Analysis16571505
Terminal-Bench-Science 0.1 (AA)Artificial Analysis28.6%9.5%
CWE-Bench (AA)Artificial Analysis–55.8%
DeepSec-Bench (AA)Artificial Analysis25.9%24.3%
CyberGym end-to-end (AA)Artificial Analysis–29.0%
FACTS SearchGoogle85.3%†–
Maths
FrontierMath Tiers 1–3Epoch AI Benchmarking Hub85.6%68.8%
FrontierMath Tier 4Epoch AI Benchmarking Hub73.2%29.3%
OTIS Mock AIMEEpoch AI Benchmarking Hub98.9%91.1%
ProofBenchVals AI99.0%49.0%
LiveBench MathematicsLiveBench95.7%87.9%†
Arena MathsArena (LMArena)15171500
Knowledge
SimpleQA VerifiedEpoch AI Benchmarking Hub59.9%41.0%
LiveBench 1.0 Data AnalysisLiveBench74.5%70.2%†
AA-OmniscienceArtificial Analysis37.114.3
MMLU-ProVals AI91.6%†86.8%
LegalBenchVals AI87.0%†84.8%
CorpFinVals AI73.2%†–
TaxEvalVals AI75.1%†72.4%
BioMysteryBenchVals AI79.3%†–
Excel Modeling BenchmarkVals AI73.6%†56.3%
Excel Modeling Benchmark v1.1Vals AI–69.1%
MedCodeVals AI63.6%†42.9%
MedScribeVals AI91.0%†88.8%
Public Benefits Bench v1.1Vals AI76.9%†68.5%
DrugDiscoveryBenchScale AI SEAL53.7%–
EBR-benchEpoch AI Benchmarking Hub45.7%–
GDP.pdfEpoch AI Benchmarking Hub24.0%–
AA-Omniscience: accuracyArtificial Analysis60.9%33.9%
AA-Omniscience: non-hallucinationArtificial Analysis39.2%70.5%
Instruction following
LiveBench LanguageLiveBench88.7%79.9%†
LiveBench Instruction FollowingLiveBench63.8%69.3%†
Arena Instruction FollowingArena (LMArena)14951478
Human preference
Arena TextArena (LMArena)14911481
EQ-Bench 4EQ-Bench1360†–
Arena Creative WritingArena (LMArena)14741456
Arena Multi-turnArena (LMArena)14871478
Arena Longer QueriesArena (LMArena)15031486
Multimodal
Arena VisionArena (LMArena)1320†–
MMMU-ProArtificial Analysis84.7%–
MMMU-Pro (Vals)Vals AI89.9%†–
MortgageTaxVals AI72.1%†–
SAGEVals AI49.4%†–
Blueprint-Bench 2Epoch AI Benchmarking Hub30.4%†–
AA-3DArtificial Analysis–783
Long context
AA-LCRArtificial Analysis79.3%79.7%
Arena DocumentArena (LMArena)1490†–
MLCRArtificial Analysis11.1%17.2%
Composite
Epoch Capabilities IndexEpoch AI Benchmarking Hub163.9†156.0†
LiveBench 1.0LiveBench80.1%76.1%†
AA Intelligence Index v4.3.2Artificial Analysis50.844.8
Vals IndexVals AI–59.7
Vals Finance IndexVals AI–60.5%
Vals RSI IndexVals AI33.0%22.9%
Vals Science IndexVals AI47.9%†–
Vals Multimodal IndexVals AI73.9%†–
Vals Router IndexVals AI–59.7%†
AA Cyber IndexArtificial Analysis–36.4
Safety & honesty
FORTRESSScale AI SEAL16.8%†28.2%
Metadata
AA Openness IndexArtificial Analysis–33.3

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

What changed

  • 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).
  • Claude Opus 5 (max): 1 new result (Vals RSI Index).
  • GLM 5.3 (max): 6 new results (Code Migration v1.1, Excel Modeling Benchmark v1.1, Finance Agent v2.1…).
  • Claude Opus 5 (max): 1 score revision.
  • GLM 5.3 (max): 1 score revision.
  • Claude Opus 5 (max): 1 new result (ForecastBench); 1 score revision.
  • GLM 5.3 (max): 4 score revisions; cached input price moved from $0.26/M to $0.30/M (15% up).
  • Claude Opus 5 (max): BenchLeader Index 68.3 → 68.9.
  • Claude Opus 5 (max): 4 new results (GDPval-AA v2.1, AA-Omniscience: accuracy, AA-Omniscience: non-hallucination…).

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

Claude Opus 5 vs GLM 5.3: questions

Is Claude Opus 5 better than GLM 5.3?
Claude Opus 5 (max) leads on quality: 68.8 vs 63.5. The BenchLeader Index combines every independent quality benchmark; Claude Opus 5 (max) is ahead overall as of 2026-10-26, but check the category scores for your use.
Which is cheaper, Claude Opus 5 or GLM 5.3, and will that change?
GLM 5.3 is cheaper today: $2.15 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 5 better than GLM 5.3 for coding?
Claude Opus 5 scores higher in coding (70 vs 62 on the category index, where 50 is average). On Arena WebDev, the top shared coding benchmark, it is 1691 for Claude Opus 5 against 1623 for GLM 5.3.
Is Claude Opus 5 better than GLM 5.3 for agents and tool use?
Claude Opus 5 scores higher in agentic tasks (64 vs 58 on the category index, where 50 is average). On Terminal-Bench, the top shared agents & tools benchmark, it is 51.8% for Claude Opus 5 against 41.8% for GLM 5.3.
Which is better for long documents?
They are level on long-context benchmarks (65 vs 65). Both accept 1,000,000 tokens of context.
Which has lower latency?
GLM 5.3 (max) answers first: 22.7 s against 68.7 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.
Which is faster, Claude Opus 5 or GLM 5.3?
GLM 5.3 streams faster: 98 against 51 output tokens per second.
Do independent benchmarks agree on Claude Opus 5 vs GLM 5.3?
Across the 43 benchmarks both have, Claude Opus 5 leads 37, GLM 5.3 leads 6. Every overall index that rates both puts Claude Opus 5 ahead.
When was GLM 5.3 released?
GLM 5.3 was released on 18 Aug 2026.