Z.AI Logo
about
Teach machines to think like humans

About us

Founded in 2019 out of Tsinghua University, Z.ai is a large-scale independent model provider in China. Our mission is to make machines think like humans through the pursuit of AGI.

Key Achievements:

- Proprietary Technology: Creators of the GLM pre-training architecture, offering superior robustness and compatibility with 40+ domestic chips.

- A Historic Breakthrough in Coding: We have achieved a critical milestone in coding capabilities. On Code Arena—the globally recognized benchmark based on over one million user blind tests—GLM tied for Global #1 in Coding alongside Anthropic and OpenAI. Surpassing overseas closed-source models like Google Gemini and xAI's Grok. GLM is the first Chinese model to achieve this ranking.

- Unified Capabilities: In 2025, we launched GLM-4.5/4.6, the first models to natively fuse reasoning, coding, and agent capabilities, ranking #1 domestically and globally among open-source models across 12 authoritative benchmarks.

- Market Dominance: Z.ai has consistently ranked in the Global Top 10 on OpenRouter for usage, with API revenue surpassing all other domestic models combined.

Trusted by 2.7 million+ developers and widely applied across finance, energy, manufacturing, and more, Z.ai has doubled its revenue for three consecutive years.
about

Development Milestones

2025

12

December

Zhipu's next-generation flagship model GLM-4.7 and China's top-performing coding model. Ranks #1 among open-source models and #4 globally on LM Arena, and #1 on OpenRouter's trending API list.

09

September

Zhipu's new flagship model GLM-4.6 achieves SOTA among global open-source models with multiple capabilities!

07

July

Launched GLM-4.5 on the Large Model Open Platform, now serving Fortune 500 companies and millions of users.

03

March

AutoGLM Rumination, the first AI Agent with deep research and operational capabilities was launched.

2024

12

December

Unveiled GLM-Zero-Preview.

10

October

Unveiled AutoGLM.

08

August

Launched GLM-4-Plus as the latest foundation model.AI video call launched in Qingyan.

01

January

Foundation model GLM-4 launched.

2023

03

March

ChatGLM released, and ChatGLM-6B generated 20 milion global downloads.

2022

09

September

GLM-130B was open sourced.

2021

03

March

GLM introduced.

2019

Founded.