Refresh Models —Deals —Regulatory —Sources —

Claude Fable 5.1 vs Quasar 438B (max, based on GLM-5.2)

14 SHARED BENCHMARKS

Across 14 shared benchmarks, Claude Fable 5.1 scores higher on 13 and Quasar 438B (max, based on GLM-5.2) on 1. The widest gap is OmniScience Accuracy, where Claude Fable 5.1 scores 67.2 against 15.5.

ANTHROPICVSZ.AI14 SHARED13–1 HEAD-TO-HEAD

BenchmarkClaude Fable 5.1MARGINQuasar 438B (max, based on GLM-5.2)

AA Agentic Index61.336.9

AA Intelligence65.743

AA-Omniscience43.5-2.6

Artificial Analysis Coding Index81.661.2

critpt31.19.4

gdpval67.740.9

GPQA Diamond93.773.2

HLE6518.7

OmniScience Accuracy67.215.5

OmniScience Non-Hallucination34.478.6

TauBench V3 - Banking47.228.7

Terminal-Bench 2.191.469.3

Best tracked score per model per benchmark (default configuration; source-attributed). ↓ marks lower-is-better metrics. Open either model for its full surface, provenance and pricing.