Z.AI · LIVEBENCH 2026-06-25

GLM-5.3

Published benchmark results for this specific model configuration.

glm-5.3

Overall score76.1 / 100

Cost / successful task$0.450USD · benchmark workload

Weight accessOpen weights

Where this configuration performs

Category averages from the same release, on a 0–100 scale.

Reasoning85.8

Coding79.0

Agentic Coding60.9

Mathematics87.9

Data Analysis70.2

Language79.9

Instruction Following69.3

Task-level results

theory of mind88.5zebra puzzle88.8spatial96.0logic with navigation70.0code generation77.5code completion80.4javascript72.7typescript50.0python60.0AMPS Hard99.0integrals with game69.0math comp96.1olympiad87.5consecutive events72.3tablejoin42.3tablereformat96.1connections95.3plot unscrambling60.2typos84.0paraphrase64.7simplify65.8story generation72.1summarize74.5

Source files: Scores ↗ Categories ↗ Costs ↗ . Missing cost is unknown, never free. Compare configurations within the same release.