# GLM-5.3 Benchmark Scores & Performance

> GLM-5.3 by Z.AI scores 65.44/100 overall, ranking #27 out of 512 AI models.

Canonical page: https://benchlm.ai/models/glm-5-3

Last updated: September 28, 2026

## Model Details

| Property | Value |
|----------|-------|
| Creator | Z.AI |
| Source Type | Open Weight |
| Reasoning Type | Reasoning |
| Context Window | 1M |
| Official model card | [Z.AI GLM-5.3 model card](https://huggingface.co/zai-org/GLM-5.3) |
| Overall Score | 65.44/100 |
| Overall Rank | #27 of 512 |

## Family & Coverage

- Family: GLM-5
- Variant: flagship (5.3)
- Benchmarks covered: 48 of 486
- Sibling models: [GLM-5.2](/models/glm-5-2), [GLM-5.3-Flash](/models/glm-5-3-flash), [GLM-5.1](/models/glm-5-1), [GLM-5](/models/glm-5), [GLM-5-Turbo](/models/glm-5-turbo), [GLM-5V-Turbo](/models/glm-5v-turbo), [GLM-5 (Reasoning)](/models/glm-5-reasoning)
- Related earlier model: [GLM-5.2](/models/glm-5-2)
- Coverage note: BenchLM currently has partial benchmark coverage for this model, so the overall score is conservative.

## Agentic Benchmarks

| Benchmark | Score |
|-----------|-------|
| [Terminal-Bench 2.1](/benchmarks/terminalbench21) | 88.2% |
| [terminalBench3](/benchmarks/terminalbench3) | 28.3% |
| [CyberGym](/benchmarks/cybergym) | 84.5% |
| [ExploitGym](/benchmarks/exploitgym) | 15.0% |
| [Toolathlon-Verified](/benchmarks/toolathlonverified) | 73.0% |
| [AutomationBench](/benchmarks/automationbench) | 48.2% |
| [Agents' Last Exam](/benchmarks/agentslastexam) | 28.5% |
| [HLE w/ tools](/benchmarks/hlewithtools) | 62.5% |
| [GDPval-AA](/benchmarks/gdpvalaa) | 1769 |
| [GDPval-AA](/benchmarks/gdpvalaanormalized) | 57.2% |
| [AA Tau3 Banking](/benchmarks/aatau3banking) | 50.3% |
| [AA Briefcase](/benchmarks/aabriefcaseelo) | 1517 |
| [AA EnterpriseOps-Gym](/benchmarks/aaenterpriseopsgym) | 36.4% |
| [Terminal-Bench 2.1 (Vals)](/benchmarks/valsterminalbench21) | 71.5% |
| [AA AutomationBench](/benchmarks/aaautomationbench) | 62.2% |
| [AA Agentic Index](/benchmarks/aaagenticindex) | 53.4% |
| [AA ITBench](/benchmarks/aaitbench) | 46.1% |
| [aaTerminalBench21](/benchmarks/aaterminalbench21) | 83.9% |
| [AA Terminal-Bench 4.0](/benchmarks/aaterminalbench4) | 41.9% |
| [GDP.pdf](/benchmarks/aagdppdf) | 11.2% |

## Coding Benchmarks

| Benchmark | Score |
|-----------|-------|
| [Terminal-Bench 2.1](/benchmarks/terminalbench21) | 88.2% |
| [terminalBench3](/benchmarks/terminalbench3) | 28.3% |
| [DeepSWE](/benchmarks/deepswe) | 66.9% |
| [NL2Repo](/benchmarks/nl2repo) | 58% |
| [ProgramBench](/benchmarks/programbench) | 19.0% |
| [FrontierSWE](/benchmarks/frontierswe) | 78.1% |
| [sweMarathon](/benchmarks/swemarathon) | 42.5% |
| [PostTrain Bench](/benchmarks/posttrainbench) | 39.8% |
| [AA-SciCode](/benchmarks/aascicode) | 59.0% |
| [VulcanBench v3](/benchmarks/vulcanbench) | 78.3% |
| [OpenHarmony Bench](/benchmarks/openharmonybench) | 60.8% |
| [FrontierSWE v2](/benchmarks/frontierswev2) | 30.2% |
| [LiveCodeBench (Vals)](/benchmarks/valslivecodebench) | 80.5% |
| [SWE-bench (Vals)](/benchmarks/valsswebench) | 95.4% |
| [AA Coding Index](/benchmarks/aacodingindex) | 74.8% |

## Multimodal & Grounded Benchmarks

| Benchmark | Score |
|-----------|-------|
| [Design Arena Website](/benchmarks/designarenawebsite) | 1309 |

## Reasoning Benchmarks

| Benchmark | Score |
|-----------|-------|
| [AA-LCR](/benchmarks/lcr) | 79.7% |
| [CritPt](/benchmarks/critpt) | 19.1% |
| [MLCR-AA](/benchmarks/aamlcr) | 48.3% |

## Knowledge Benchmarks

| Benchmark | Score |
|-----------|-------|
| [Artificial Analysis Intelligence Index](/benchmarks/artificialanalysis) | 44.8% |
| [AA-GPQA Diamond](/benchmarks/aagpqadiamond) | 91.7% |
| [AA-HLE](/benchmarks/aahle) | 42.3% |
| [AA-Omniscience Index](/benchmarks/aaomniscienceindex) | 14.3% |
| [AA-Omniscience Accuracy](/benchmarks/omniscienceaccuracy) | 33.9% |
| [AA-Omniscience Hallucination Rate](/benchmarks/omnisciencehallucinationrate) | 29.6% |
| [GPQA Diamond (Vals)](/benchmarks/valsgpqadiamond) | 88.1% |
| [MMLU-Pro (Vals)](/benchmarks/valsmmlupro) | 86.8% |

## Other Z.AI Models

- [GLM-5.2](/models/glm-5-2) - Score: 62.62
- [GLM-5.3-Flash](/models/glm-5-3-flash) - Score: 60.62
- [GLM-5.1](/models/glm-5-1) - Score: 57.11
- [GLM-5](/models/glm-5) - Score: 54.39
- [GLM-5-Turbo](/models/glm-5-turbo) - Score: 52.66
- [GLM-4.7](/models/glm-4-7) - Score: 49.19
- [GLM-5V-Turbo](/models/glm-5v-turbo) - Score: 48.84
- [GLM-4.6](/models/glm-4-6) - Score: 39.76
- [GLM-4.7-Flash](/models/glm-4-7-flash) - Score: 38.08
- [GLM-4.5-Air](/models/glm-4-5-air) - Score: 34.63
- [GLM-4.5](/models/glm-4-5) - Score: not computed
- [GLM-5 (Reasoning)](/models/glm-5-reasoning) - Score: not computed