Zhipu AI Launches New Model Better at Coding, Learning
00:00/00:00
Listen to this article
1x
Subscribe now
Listen to the full version
Listen to the full version
Chinese artificial intelligence startup Zhipu AI on Thursday officially released a new open-source large language model, GLM-5, designed for complex system engineering and autonomous agent tasks.
The GLM-5 model closely matches the real-world programming performance of Claude Opus 4.5, an AI model launched by U.S. firm Anthropic in November 2025, Beijing-based Zhipu AI said.
You've accessed an article available only to subscribers
VIEW OPTIONS
Unlock exclusive discounts with a Caixin group subscription — ideal for teams and organizations.
Subscribe to both Caixin Global and The Wall Street Journal — for the price of one.
- RELATED
Share this article
Open WeChat and scan the QR code
- DIGEST HUB
Digest Hub
Back
Explore the story in 30 seconds
- Zhipu AI released GLM-5, an open-source large language model with 744 billion parameters, ranking fourth globally in comprehensive capabilities.
- GLM-5 outperforms Google’s Gemini 3 Pro and Kimi K2.5, offers enhanced coding and autonomous agent functions, and supports domestic Chinese chips.
- After raising $534 million in its 2025 Hong Kong IPO, Zhipu AI reported a $341 million net loss in H1 2025, with R&D expenses nearly doubling.
AI generated, for reference only
Who’s Who
- Zhipu AI
- Zhipu AI, a Chinese AI startup, recently released GLM-5, an open-source large language model excelling in complex system engineering and autonomous agent tasks. It ranks fourth globally in comprehensive capabilities, outperforming some rivals and matching Claude Opus 4.5 in real-world programming. GLM-5 uses a Mixture-of-Experts architecture with 744 billion parameters and is optimized for domestic chipmakers, crucial for the Chinese market. The company went public in January, raising significant capital for AI research despite reporting substantial net losses in the first half of 2025.
- Anthropic
- Anthropic is a U.S. firm that launched the AI model Claude Opus 4.5 in November 2025. This model's real-world programming performance is closely matched by Zhipu AI's new open-source large language model, GLM-5. In comprehensive capabilities, Claude Opus 4.5 is listed among the top three models globally, trailing only Claude Opus 4.6 and GPT-5.2.
- Moonshot AI
- Moonshot AI (月之暗面) is a Chinese artificial intelligence company that developed the Kimi K2.5 model. Its Kimi K2.5 model was outperformed by Zhipu AI's newly released GLM-5 model.
- DeepSeek
- DeepSeek launched the "DeepSeek Sparse Attention" mechanism in September 2025. This mechanism is adopted by Zhipu AI's GLM-5 model to improve learning capabilities from long-range interactions and reduce deployment costs.
- Huawei Technologies Co. Ltd.
- Huawei Technologies Co. Ltd. is a domestic Chinese chipmaker. Zhipu AI’s new GLM-5 large language model has completed deep inference adaptation with Huawei, ensuring the model's stable operation on Chinese-made chip clusters. This highlights Huawei's role in supporting advanced AI model infrastructure within China.
- Moore Threads Technology Co. Ltd.
- Moore Threads Technology Co. Ltd. is a Chinese chipmaker. Zhipu AI's new GLM-5 model has completed deep inference adaptation with Moore Threads, ensuring the model's stable operation on their domestically produced chip clusters.
- Cambricon Technologies Corp. Ltd.
- Cambricon Technologies Corp. Ltd. is a Chinese chipmaker. Zhipu AI's new large language model, GLM-5, has completed deep inference adaptation with Cambricon, ensuring the model's stable operation on home-made chip clusters within the Chinese market. This collaboration highlights efforts to integrate advanced AI models with domestic hardware in China.
AI generated, for reference only
What Happened When
- First half of 2025：
- Zhipu AI reported a net loss of 2.36 billion yuan and research and development expenses nearly doubled to 1.6 billion yuan.
- September 2025：
- DeepSeek launched the 'DeepSeek Sparse Attention' mechanism.
- November 2025：
- U.S. firm Anthropic launched AI model Claude Opus 4.5.
- January 8, 2026：
- Zhipu AI held its IPO on the Hong Kong Stock Exchange, raising HK$4.17 billion.
- February 12, 2026：
- Zhipu AI officially released the GLM-5 open-source large language model.
AI generated, for reference only
- MOST POPULAR
Sign in with Google
Sign in with Facebook
Sign in with 财新