GLM-5.3

GLM-5.3 is a reasoning model from Z.AI in the GLM-5 family. 51 benchmarks count toward its score, in 6 categories.

availableShows if the model has enough results for an index.
IndexOverall score. 50 is the middle.69.6 ±3.8
CoverageShare of the index weight with results.85%
SpeedOutput tokens per second.152/s
Input / 1MUS dollars per 1M input tokens.$0.84 batch $0.7
Output / 1MUS dollars per 1M output tokens.$2.64 batch $2.2 US dollars per 1M output tokens in a batch.
ContextMaximum tokens in one request.1.31M
EloLMArena rating and rank.1475 (#20)

50 is the middle of the board. The range shows the doubt in the index. Batch work costs less.

10,960 votes. Elo shows what people prefer. It does not change the score.

CapabilitiesScore per category. 50 is the middle.

50 is the middle
AgenticMulti-step tasks with tools.
75.2
CodingCode writing and repair.
67.5
ReasoningLogic problems and puzzles.
67.6
MultimodalTasks with images and text.
N/A
KnowledgeFacts and expert knowledge.
64.0
MultilingualTasks in many languages.
N/A
InstructionTasks with strict rules in the prompt.
69.4
MathMath problems.
65.8

Results

51 counted
BenchmarkThe test name.CategoryThe capability that the test measures.ResultThe score from the publisher.IndexThis result on the index scale.RunThe settings of the run.DateDate of the result.Published byThe source of the result.
SWE-benchCoding95.4%74.11 Sept 2026Vals AI
Artificial Analysis GPQA DiamondKnowledge91.7%62.5Artificial Analysis
OTIS Mock AIME 2024-2025Math91.1%62.5max effortEpoch AI
GPQA diamondKnowledge90.9%62.2max effortEpoch AI
Terminal-Bench 2.1 (provider run)Agentic88.2%76.1DeepSeek-AI
Terminal-Bench 2.1 (provider run)Agentic88.2%76.1DeepSeek-AI
GPQA DiamondKnowledge88.1%59.61 Sept 2026Vals AI
LiveBench MathematicsMath87.9%65.325 Jun 2026LiveBench
MMLU ProKnowledge86.8%57.31 Sept 2026Vals AI
LiveBench ReasoningReasoning85.8%75.225 Jun 2026LiveBench
CyberGymAgentic84.5%74.0Zhun Wang et al.
LiveCodeBenchCoding80.5%57.51 Sept 2026Vals AI
LiveBench LanguageKnowledge79.9%66.925 Jun 2026LiveBench
Artificial Analysis Long Context ReasoningReasoning79.7%63.4Artificial Analysis
LiveBench CodingCoding79.0%68.725 Jun 2026LiveBench
VulcanBench v3Coding78.3%58.6VulcanBench contributors
Vibe Code Bench v1.1Coding78.1%74.7OpenHands21 Sept 2026Vals AI
FrontierSWECoding78.1%Evan Chu et al.
Artificial Analysis Coding IndexCoding74.8%71.6Artificial Analysis
Toolathlon-VerifiedAgentic73.0%71.4Moonshot AI
Terminal-Bench 2.1Agentic71.5%66.321 Sept 2026Vals AI
LiveBench Data AnalysisReasoning70.2%53.525 Jun 2026LiveBench
LiveBench Instruction FollowingInstruction69.3%69.425 Jun 2026LiveBench
FrontierMath-Tiers-1-3-v2-PrivateMath68.8%70.2max effortEpoch AI
IOICoding68.4%75.521 Sept 2026Vals AI
DeepSWEAgentic66.9%72.4Datacurve AI
Humanity's Last Exam with toolsAgentic62.5%75.1DeepSeek-AI
Artificial Analysis AutomationBenchAgentic62.2%74.9Artificial Analysis
LiveBench Agentic CodingAgentic60.9%74.025 Jun 2026LiveBench
OpenHarmony Bench v1.0Coding60.8%70.0OpenHarmony Bench authors
Artificial Analysis SciCodeCoding59.0%74.7Artificial Analysis
NL2RepoCoding58.0%70.6MiniMax
GDPval-AA normalizedAgentic57.3%79.6Artificial Analysis
Artificial Analysis Agentic IndexAgentic53.4%80.7Artificial Analysis
Artificial Analysis Tau3-BankingAgentic50.3%80.9Artificial Analysis
ProofBench v1.1Math49.0%69.221 Sept 2026Vals AI
Medical Long Context Reasoning (MLCR-AA)Reasoning48.3%80.0Wisedocs and Artificial Analysis
AutomationBenchAgentic48.2%95.0Moonshot AI
SkillsBenchCoding47.5%63.3OpenHands11 Sept 2026Vals AI
Artificial Analysis Intelligence IndexKnowledge44.8%78.4Artificial Analysis
Code MigrationCoding44.2%73.221 Sept 2026Vals AI
SWE-MarathonCoding42.5%Abundant AI and BenchFlow
Artificial Analysis Humanity's Last ExamKnowledge42.3%70.4Artificial Analysis
Terminal-Bench 4.0.0Agentic41.8%88.9max effort · Claude Code21 Sept 2026Terminal-Bench
SimpleQA VerifiedKnowledge41.0%59.5max effortEpoch AI
PostTrain BenchCoding39.8%Moonshot AI
Artificial Analysis EnterpriseOps-GymAgentic36.4%57.2Artificial Analysis
Artificial Analysis Omniscience AccuracyKnowledge33.9%55.8Artificial Analysis
Mystery Game PuzzlesReasoning33.0%69.0max effortEpoch AI
FrontierSWE v2Coding30.2%71.2Proximal
FrontierMath-Tier-4-v2-PrivateMath29.3%62.0max effortEpoch AI
Agents' Last ExamAgentic28.5%68.0DeepSeek-AI
terminalBench3Agentic28.3%Benchmark authors
terminalBench3Agentic28.3%Benchmark authors
Terminal-Bench 4.0Agentic25.3%77.221 Sept 2026Vals AI
Chess PuzzlesReasoning21.0%50.7max effortEpoch AI
Vibe Code Bench 1-100Coding20.0%73.8OpenHands16 Sept 2026Vals AI
Critical Physics TasksReasoning19.1%78.4Artificial Analysis
ProgramBench: Can Language Models Rebuild Programs From Scratch?Coding19.0%34.5John Yang et al.
ExploitGymAgentic15.0%72.0Zhun Wang et al.
Agent Arena task outcomeAgentic8.576.9max effort15 Sept 2026LMArena
Agent Arena steerabilityAgentic1.969.5max effort15 Sept 2026LMArena
ProgramBenchCoding1.5%21 Sept 2026Vals AI
Agent Arena command recoveryAgentic-1.066.2max effort15 Sept 2026LMArena

51 benchmarks count, from 58 of 64 results. A grey row does not count. Too few models took that benchmark.

Sources

BenchLM benchmark aggregationUsed with attribution; per-benchmark results credited to their original authorsOpenRouter, collected directlyNo licence statedVals AI, collected directlyNo licence stated. Read from the public leaderboard and credited to Vals AIEpoch AI, collected directlyCC BY — free to use and redistribute with attributionLiveBench, collected directlyNo licence stated for the leaderboard. The site repo serving these CSVs has no LICENSE; the harness repo carries upstream Apache-2.0 and MIT copies that cover code, not resultsTerminal-Bench, collected directlyNo licence stated for the leaderboard. The harness repo is Apache-2.0LMArena, collected directlyCC BY 4.0 (lmarena-ai/leaderboard-dataset on Hugging Face)

Same level, lower price

GPT-6 Luna68.4 · $0.5DeepSeek V4.1 Flash70.6 · $0.6MiMo-V2.6-Pro73.5 · $0.87

More from Z.AI

GLM-5.3-Flash65.6GLM-5.262.7GLM-5.157.8GLM-552.2GLM-4.750.2