Compare AI models

Specifications
Parameters
>1T744B
Context window
256k1M
API pricing
Input price
—$1.40
Output price
—$4.40
Cached input price
—$0.26
Cheapest input
—$0.007Wafer
Cheapest output
—$1.44DeepInfra
—$182Relace · fp4
Benchmarks
Humanity's Last Exam · text only, no tools
30.2%40.5%
Humanity's Last Exam · with tools
49.8%54.7%
GPQA Diamond
87.4%91.2%
Overview
CompanyQwenZ.ai
Release dateJan 26 2026Jun 16 2026
AccessClosedOpen Weight
Model detailsView modelView model

Frequently asked questions

GLM-5.2 leads Qwen3-Max-Thinking on 3 of the 3 benchmarks they both report (Humanity's Last Exam, GPQA Diamond). Only GLM-5.2 has a verified first-party price for an available API: $1.40 per million input tokens and $4.40 per million output tokens. No currently available pay-as-you-go API rate is tracked for Qwen3-Max-Thinking. Qwen3-Max-Thinking shipped 141 days before GLM-5.2, so benchmark comparisons should account for the intervening progress.

Qwen3-Max-Thinking has >1T parameters, while GLM-5.2 has 744B. Context windows are 256k (Qwen3-Max-Thinking) vs 1M (GLM-5.2). Qwen3-Max-Thinking is closed, while GLM-5.2 is open weight.

On Humanity's Last Exam · text only, no tools, GLM-5.2 leads at 40.5% vs Qwen3-Max-Thinking at 30.2%. On Humanity's Last Exam · with tools, GLM-5.2 leads at 54.7% vs Qwen3-Max-Thinking at 49.8%. On GPQA Diamond, GLM-5.2 leads at 91.2% vs Qwen3-Max-Thinking at 87.4%.