GLM-5.3 benchmark ranking on every AI leaderboard

Name: GLM-5.3
Author: Z.ai (Zhipu)

open weights · #21 overall, podium score 8.0 · top 10 on 4 of 11 · released Aug 14, 2026

11 of 11 boards

Rank	Board	Category	Score	Of
5	MCP AtlasAgents and tools · as “GLM 5.3”	Agents and tools	84.2%	34
7	SWE-Bench ProCoding · as “GLM-5.3 (mini-swe-agent) max”	Coding	84.3%	11
9	AA IntelligenceReasoning · as “GLM-5.3 (max)”	Reasoning	44.8	100
9	OpenRouterUsage · as “z-ai/glm-5.3-20260816”	Usage	3.0T tok	19
14	WebDev ArenaCoding · as “glm-5.3-max”	Coding	1,622	100
15	AA SpeedSpeed · as “GLM-5.3 (max)”	Speed	57 tok/s	24
15	AA PricePrice · as “GLM-5.3 (max)”	Price	$2.15	27
16	Text ArenaChat · as “glm-5.3-max”	Chat	1,483	100
22	Epoch ECIReasoning · as “GLM-5.3”	Reasoning	155.6	100
25	GPQA DiamondReasoning · as “glm-5.3_max”	Reasoning	90.9%	100
28	LiveBenchReasoning · as “glm-5.3”	Reasoning	76.1%	59

Category ranks and more models

Scores as published by each board on the capture date. Model names and logos belong to their owners; logos via logo.dev.