Skip to main content

Z.ai

Category: Foundation Models / LLMs

Z.ai (formerly Zhipu AI) is a Chinese AI company behind the GLM series of large language models, one of China's "AI tigers" and the first Chinese foundation model company to go public via IPO on the Hong Kong Stock Exchange. Z.ai was founded in 2019. The company is led by Zhang Peng (张鹏). Based in Beijing, China. Total funding raised: $1.5B. Latest round: IPO. Key investors include Alibaba Group, Tencent, Meituan, Ant Group, Xiaomi, HongShan (Sequoia China), Qiming Venture Partners, Prosperity7 Ventures (Saudi Aramco), Hangzhou City Investment Group, Shanghai Chengtou, Pudong Innovation, Zhangjiang Group.

AMW Analysis

Z.ai (formerly Zhipu AI), a Beijing-based developer of the GLM foundation model family, has continued rapid product releases through mid-2026 following its IPO on the Hong Kong Stock Exchange, backed by a large domestic investor base including Alibaba, Tencent, Ant Group, and Xiaomi. In April, the company released GLM-5.1, a 754B-parameter Mixture-of-Experts model under an MIT license that reported top scores on SWE-Bench Pro, supporting long autonomous coding sessions. This was followed in July by GLM-5.2, adding agentic capabilities, a one-million-token context window, and open-weight access aimed at software engineering and enterprise automation workflows optimized for domestic compute. Alongside the model launch, Z.ai introduced ZCode, a free coding assistant built on GLM-5.2, positioning it against established coding-assistant products. Separate reporting indicates Z.ai is pursuing in-house inference chip development to reduce dependence on external chip suppliers, a move linked to discussion of potential restrictions on overseas models in China. Together, the news flow shows a company advancing its open-weight model lineup, extending into developer tooling, and addressing supply-chain independence for compute.

AMW analysis, generated from 4 tracked news signals.

Founded
2019
Headquarters
Beijing, China
Total funding
$1.5B

Value proposition

Developing safe and beneficial AGI through the GLM model family, offering open-source foundation models, API services, and on-premise AI solutions for enterprises and governments, with a focus on Chinese domestic AI ecosystem independence.

Products and solutions

GLM-5.2 (flagship LLM with 1M context window), GLM-4.5/4.6/4.7 series, ChatGLM (conversational AI), GLM-4-Voice (emotional voice model), AutoGLM (AI agent), GLM-PC (autonomous PC agent), Ying (text-to-video), GLM-Image (image generation), CodeGeeX (code generation), AMiner (academic knowledge graph)

Unique value

China's first publicly listed AGI foundation model company with a full-stack AI portfolio (language, vision, video, voice, agents) and deep integration with domestic semiconductor ecosystem (Huawei Ascend, Cambricon, Moore Threads).

Target customer

Chinese government agencies; state-owned enterprises; enterprise customers requiring on-premise AI deployment; global developers via open-source models and API platform

Industries served

Government & Public Sector, Enterprise IT, Education, Healthcare, Finance, Smart City

Technology advantage

GLM (General Language Model) architecture using autoregressive blank infilling; full-stack AI capabilities (text, vision, video, voice, agents); open-source MIT-licensed models; compatibility with domestic Chinese chips (Huawei Ascend, Cambricon, Moore Threads); 745B-parameter GLM-5 model trained on Huawei Ascend chips

How they differentiate

Strong academic lineage from Tsinghua University's Knowledge Engineering Group (KEG); early and aggressive open-source strategy (MIT License since July 2025); deep partnerships with domestic chipmakers (Huawei Ascend, Cambricon) enabling AI sovereignty; first-mover in Chinese AI IPO; strong government/enterprise on-premise deployment model (85% of 2024 revenue).

Main competitors

DeepSeek, Moonshot AI (Kimi), MiniMax, Baichuan AI, 01.AI, StepFun

Key partnerships

Huawei (Ascend chip collaboration, GLM-Image co-development), Cambricon Technologies, Moore Threads, Tsinghua University KEG Lab, Samsung China (strategic AI phone partnership), Intel (AI PC ecosystem), Honor (AI joint lab), ASUS

Notable customers

China Association for Science and Technology, Chinese government agencies (Hangzhou city projects), various state-owned enterprises, Southeast Asian and Middle Eastern governments

Major milestones

2019: Founded as spinout from Tsinghua University, 2023: Raised ¥2.5B from Alibaba, Tencent, Meituan, Ant Group, Xiaomi, 2024: Reached $2.8B valuation, 2025: Rebranded as Z.ai internationally, Jan 2026: IPO on HKEX (2513.HK), raised HK$4.35B, Feb 2026: Released GLM-5 (745B parameters), Jun 2026: Released GLM-5.2 with 1M context window

Growth metrics

2024 revenue: ~$37.8M; 2025 revenue: ~$106.9M (132% YoY growth); 2025 net loss: ¥4.72B; Market cap peaked at ~HK$880B (~$112B) in May 2026

Market positioning

One of China's "AI tigers" and the third-largest LLM market player in China's AI industry per IDC. The first Chinese foundation model company to go public (HKEX: 2513). Known for open-source GLM models and strong government/enterprise revenue.

Geographic focus

China (primary); Southeast Asia; Middle East; Africa

Patents and IP

GLM architecture patent (autoregressive blank infilling); multiple model releases under MIT open-source license; AMiner academic knowledge graph platform

About Zhang Peng (张鹏)

PhD from Tsinghua University Department of Computer Science and Technology; 2018 Innovation Leader Project of Tsinghua University; Researcher at Zhipu.AI (aminer.cn) since 2020

Latest news about Z.ai

More Foundation Models / LLMs companies

Official website: