evals.report
BenchmarksLabsCompareRun guidesIn the wild
Z.aiGLM

GLM-5

Z.ai · GLM. Released Feb 11, 2026.

GLM-5 is a model from Z.ai in the GLM family, released Feb 11, 2026. evals.report tracks 25 reported GLM-5 benchmark scores across ARC-AGI-1, GPQA Diamond, SWE-bench Verified, FrontierMath, AIME (OTIS Mock), ARC-AGI-2, LMArena, WeirdML, and 17 more — each shown with its benchmark, metric, source status, and date, and never combined into a single ranking.

Open25 results

Benchmark results 25

Compare this model
BenchmarkCategoryScoreMetricStatusDate
ARC-AGI-1Reasoning44.67%accuracyOfficialFeb 11, 2026Details
GPQA DiamondReasoning87.8%accuracyOfficialFeb 11, 2026Details
SWE-bench VerifiedCoding72.1%% resolvedOfficialFeb 11, 2026Details
FrontierMathReasoning16.43%accuracyOfficialFeb 11, 2026Details
AIME (OTIS Mock)Reasoning80.0%accuracyOfficialFeb 11, 2026Details
ARC-AGI-2Reasoning4.86%accuracyOfficialFeb 11, 2026Details
LMArenaChat preference1445source-defined ratingOfficialFeb 11, 2026Details
WeirdMLCoding48.2%average accuracyOfficialFeb 11, 2026Details
PostTrainBenchAgents13.88%weighted average scoreOfficialFeb 11, 2026Details
Artificial Analysis Intelligence IndexReasoning49.8IndexUnverifiedFeb 11, 2026Details
Epoch Capabilities IndexReasoning146.6IndexOfficialFeb 11, 2026Details
SWE-rebenchCoding62.8%Resolved rate (pass@1)OfficialFeb 11, 2026Details
GAIA: A Benchmark for General AI AssistantsAgents33.8%accuracyUnverifiedFeb 11, 2026Details
AIME 2026Reasoning95.83%accuracyOfficialFeb 11, 2026Details
GDPvalAgents1395EloOfficialFeb 11, 2026Details
SciCodeCoding46.2%accuracyUnverifiedFeb 11, 2026Details
Global-MMLUReasoning81.9%accuracyUnverifiedFeb 11, 2026Details
WebDev ArenaChat preference1436EloVerifiedFeb 11, 2026Details
EQ-Bench Creative Writing v3Chat preference1658EloVerifiedFeb 11, 2026Details
Design ArenaChat preference1300EloVerifiedFeb 11, 2026Details
SWE-bench MultilingualCoding69.7%% resolvedOfficialFeb 11, 2026Details
MathArena HMMT February 2026Reasoning86.36%accuracyOfficialFeb 11, 2026Details
FrontierMath Tier 4Reasoning2.1%accuracyOfficialFeb 11, 2026Details
Vectara Hallucination LeaderboardOther10.1%Hallucination RateOfficialFeb 11, 2026Details
Vibe Code BenchCoding23.36%Overall accuracyVerifiedFeb 11, 2026Details