GLM-4.5 (Reasoning)

← AI Models
Unknown
MIT
355B params
Modality:
Intelligence
—
Coding
—
Math
—
Speed
—
Pricing
—

How GLM-4.5 (Reasoning) Compares

Benchmark Profile

All Benchmark Scores (17)

BenchmarkScore
GPQA79.1%
HLE14.4%
SWE-Bench Verified64.2%
AA-Index67.7%
AIME 202491%
BFCL-v377.8%
BrowseComp26.4%
Humanity's Last Exam14.4%
LiveCodeBench72.9%
MATH-50098.2%
MMLU-Pro84.6%
SciCode41.7%
TAU-bench Airline60.4%
TAU-bench Retail79.7%
Terminal-Bench37.5%
Arena Chat21.4
Arena Coding7.4

Data: Artificial Analysis · Updated: April 10, 2026