evals.report
BenchmarksLabsCompareRun guidesIn the wild
Z.aiGLM

GLM-5.3

Z.ai · GLM. Released Aug 14, 2026.

GLM-5.3 is a model from Z.ai in the GLM family, released Aug 14, 2026. evals.report tracks 5 reported GLM-5.3 benchmark scores across FrontierCode 2.0, AA-Omniscience v1.1: Knowledge and Hallucination Benchmark, FrontierCode, AA-Omniscience: Knowledge and Hallucination Benchmark, DeepSWE — each shown with its benchmark, metric, source status, and date, and never combined into a single ranking.

Open5 results

Benchmark results 5

Compare this model
BenchmarkCategoryScoreMetricStatusDate
FrontierCode 2.0Coding20.2%weighted score (Main)OfficialAug 14, 2026Details
AA-Omniscience v1.1: Knowledge and Hallucination BenchmarkReasoning6AA-Omniscience Net ScoreOfficialAug 14, 2026Details
FrontierCodeCoding40.1%weighted score (Main)OfficialAug 14, 2026Details
AA-Omniscience: Knowledge and Hallucination BenchmarkReasoning14AA-Omniscience IndexOfficialAug 14, 2026Details
DeepSWECoding66.9%% resolvedVerifiedAug 14, 2026Details