TLDR: Chinese AI firm Zhipu AI, a prominent competitor to OpenAI, has unveiled its new open-source model, GLM-4.5. This release signifies a significant advancement in China’s AI capabilities and further intensifies the competitive landscape in the field of large language models, with Zhipu aiming to challenge Western AI leaders.
Zhipu AI, a leading Chinese artificial intelligence firm and a spin-off from Tsinghua University, has officially launched its latest open-source language model, GLM-4.5. This strategic move is set to reinforce Zhipu’s position as a formidable competitor to global AI powerhouses like OpenAI, as the company continues to expand its GLM (General Language Model) series with a focus on making advanced AI more accessible.
The GLM-4.5 model is an upgrade to the firm’s existing flagship model and is designed to enhance Zhipu’s competitive standing in the global AI market. The release aligns with a broader trend within China’s AI sector, where numerous firms are increasingly adopting open-source strategies to drive adoption and influence global AI standards. Other Chinese AI companies, such as Moonshot and StepFun, have also recently introduced their own open-source models, including Moonshot’s Kimi K2 and StepFun’s non-proprietary reasoning system.
Beyond GLM-4.5, Zhipu AI has been actively releasing other advanced models. Recent open-sourced versions of its GLM-4 series and GLM-Z1 inference models include 9B and 32B parameter variants, all released under the permissive MIT license to support commercial use and offer developers greater freedom. In July 2025, Zhipu AI also launched GLM-4.1V-Thinking, an open-source multimodal reasoning model optimized for complex cognitive tasks, capable of handling multimodal inputs like images, videos, and documents with improved performance. This model is reported to perform comparably to, and in some cases surpass, larger models like OpenAI’s GPT-4o on various benchmarks.
Further demonstrating its commitment to autonomous AI systems, Zhipu AI launched AutoGLM Rumination in March 2025. This free AI agent, powered by GLM-Z1-Air and GLM-4-Air-0414 models, combines research and task execution capabilities, enabling deep research, web searches, travel planning, and report writing. In October 2024, the company also released GLM-4-Voice, an open-source end-to-end speech large language model that understands and generates Chinese and English speech with flexible adjustments for emotion, tone, speed, and dialect, supporting real-time interaction.
To further foster the broader AI industry ecosystem, Zhipu AI has established an open-source fund, committing a thousand AI chip cards and 10 million yuan (approximately USD 1.39 million) in cash to support open-source projects. The company has attracted significant investment from major Chinese tech firms, including Alibaba Group Holding and Tencent Holdings. Zhipu has also been actively involved in the ‘Special Migration Program’ for OpenAI API users following OpenAI’s API block in some regions.
Also Read:
- China’s AI Sector Forges Alliances to Counter US Tech Restrictions
- OpenAI’s GPT-5 Set for August 2025 Launch, Promising Advanced Reasoning and Expanded Context
In related news, Zhipu is reportedly re-evaluating its listing strategy, considering a potential shift of its planned initial public offering (IPO) from mainland China to Hong Kong, which could raise up to $300 million.
