Skip to main content

Zhipu AI

Category: Foundation Models / LLMs

A leading AGI infrastructure provider (rebranded internationally as Z.ai) developing the GLM foundation model series, autonomous agents, and the 'Zhipu Qingyan' consumer super-app. Zhipu AI was founded in 2019. The company is led by Zhang Peng. Based in Beijing, China. Team size: 800-900. Total funding raised: $2.0B. Latest round: IPO (Jan 2026). Key investors include Alibaba, Tencent, Prosperity7 Ventures (Aramco), Meituan, Beijing AI Industry Fund.

Founded
2019
Headquarters
Beijing, China
Team size
800-900
Total funding
$2.0B

Value proposition

Provides high-performance, cost-efficient bilingual (Chinese-English) foundation models with a unique architecture that outperforms competitors in agentic tasks and local deployment.

Products and solutions

GLM-5.2 / GLM-5.1 / GLM-5 (Flagship LLMs with MoE architecture, released Feb-Jun 2026), AutoGLM (Autonomous Agent System with 'Rumination' capabilities), CogVideoX (Open-source Text-to-Video Generation), CodeGeeX (AI Coding Assistant), Zhipu Qingyan (Consumer Generative AI App), MaaS Open Platform (bigmodel.cn)

Unique value

Pioneered the 'GLM' (General Language Model) architecture, which uses autoregressive blank-filling (unlike standard GPT decoder-only models) to optimize for both understanding and generation tasks.

Target customer

Enterprise developers (MaaS), Government/Public Sector, Consumer (via Qingyan App), and Academic Researchers

Industries served

Artificial Intelligence, Enterprise Software (SaaS), Consumer Internet, Education & Research, Media & Entertainment

Technology advantage

First Chinese AI lab to achieve 'System 2' reasoning capabilities (via AutoGLM Rumination) and successfully commercialize a full stack: from proprietary pre-training framework to a mass-market consumer app.

How they differentiate

Proprietary GLM architecture optimized for Agentic AI (AutoGLM); 'Z.ai' global brand with dual-engine strategy (Open Source + Commercial MaaS); Strongest domestic hardware adaptation (Huawei Ascend/Moore Threads).

Main competitors

MiniMax, Moonshot AI (Kimi), Baidu (Ernie Bot)

Key partnerships

Tsinghua University (Knowledge Engineering Group - KEG), Alibaba Cloud & Tencent (Cloud Infrastructure & Strategic Investment), Saudi Aramco / Prosperity7 Ventures (International Expansion), CCF (China Computer Federation)

Notable customers

Samsung (Galaxy S25), WPS (Kingsoft), Liepin, PCauto

Major milestones

Listed on HKEX (Stock Code: 02513) in Jan 2026, Rebranded to Z.ai for global expansion in 2025, Secured $400M from Prosperity7 (Saudi Aramco) in June 2024, Released GLM-5 flagship model in Feb 2026, Released GLM-5.1 in Mar 2026 and open-sourced in Apr 2026, Released GLM-5.2 with 1M-token context window in Jun 2026

Growth metrics

2025 Revenue ~$105M (724M RMB), up 132% YoY; 12,000+ Enterprise Clients; 2.9M MaaS Developers; Net loss RMB 4.7B.

Market positioning

China's leading 'AI Tiger' and Enterprise MaaS Utility; First Gen-AI unicorn to IPO.

Geographic focus

China (Mainland), Hong Kong, Middle East & SE Asia

Patents and IP

Extensive IP portfolio in pre-training algorithms (GLM), knowledge graphs (AMiner integration), and multimodal generation (CogView/CogVideo)

About Zhang Peng

Zhang Peng holds a Ph.D. from the Department of Computer Science and Technology at Tsinghua University. He is an expert in knowledge engineering and large-scale pre-training models. Prior to founding Zhipu AI, he was a key researcher at Tsinghua's Knowledge Engineering Group (KEG). He has led the development of the ChatGLM series and successfully guided the company to its IPO on the Hong Kong Stock Exchange.

Official website: