GLM 5.2

Über dieses Modell

ZAI GLM 5.2 is a 744B parameter Mixture-of-Experts (MoE) language model built on the GLM-MoE DSA architecture. It features a solid 1‑million token context window and multiple thinking effort levels for flexible reasoning and coding performance. The model achieves strong benchmark results, e.g., 54.7 on HLE (with tools), 99.2 on AIME 2026, and 91.2 on GPQA‑Diamond. It also excels in agentic tasks such as MCP‑Atlas (76.8) and Tool‑Decathlon (48.2). GLM 5.2 is released under the MIT license.

Technische Daten

Fähigkeiten

Eingabe-Modalitäten

Ausgabe-Modalitäten

Reasoning

Hybrid Standard on

Knowledge horizon