China’s new AI model GLM-4.5 threatens DeepSeek – will it also threaten OpenAI?

China's AI

In a bold move reshaping the global AI landscape, Chinese startup Z.ai has launched GLM-4.5, an open-source model touted as cheaper, smaller, and more efficient than rivals like DeepSeek.

The announcement, made at the World Artificial Intelligence Conference in Shanghai, has sent ripples across the tech sector.

What sets GLM-4.5 apart is its lean architecture. Requiring just eight Nvidia H20 chips—custom-built to comply with U.S. export restrictions—it slashes operating costs dramatically.

By comparison, DeepSeek’s model demands nearly double the compute power, making GLM-4.5 a tantalising alternative for cost-conscious developers and enterprises.

But the savings don’t stop there. Z.ai revealed that it will charge just $0.11 per million input tokens and $0.28 per million output tokens. In contrast, DeepSeek R1 costs $0.14 for input and a hefty $2.19 for output, putting Z.ai firmly in the affordability lead.

Functionally, GLM-4.5 leverages ‘agentic’ AI—meaning it can deconstruct tasks into subtasks autonomously, delivering more accurate results with minimal human intervention.

This approach marks a shift from traditional logic-based models and promises smarter integration into coding, design, and editorial workflows.

Z.ai, formerly known as Zhipu, boasts an impressive funding roster including Alibaba, Tencent, and state-backed municipal tech funds.

With IPO ambitions on the horizon, its momentum mirrors China’s broader push to dominate the next wave of AI innovation.

While the U.S. has placed Z.ai on its entity list, stifling some Western partnerships, the firm insists it has adequate computing resources to scale.

As AI becomes a battleground for technological and geopolitical influence, GLM-4.5 may prove to be a powerful competitor.

But it has some way yet to go.

China’s tech stocks rally to 13-month high on new stimulus

Tech stocks up China

Chinese technology stocks, such as the previously underperforming Alibaba, have surged this week, reaching peaks not observed in over a year

The stock surge follows the announcement of stimulus measures by China’s central bank to boost the world’s second-largest economy.

On Thursday 26th September 2024 in the U.S., Alibaba’s shares closed above $100 for the first time since August 2023.

Tencent’s shares ended at their highest point in over two and a half years.

The AI Race between China and the U.S.

AI development in China and U.S.

Artificial Intelligence (AI) has become a pivotal battleground in the technological race between China and the United States.

“AI is expected to become a crucial component of economic and military power in the near future,” Stanford University’s Artificial Intelligence Index Report 2023 stated.

Both countries are significantly investing in AI research and development, striving to achieve a leading role in this revolutionary sector. This post looks at the major figures in China’s AI scene, their progress, and their comparison with their American counterparts.

China’s AI Landscape

China’s AI aspirations are propelled by a number of significant technology firms, each forging their own AI models and applications.

Baidu: Often referred to as the ‘Google of China,’ Baidu leads in AI development. Its premier AI model, ERNIE (Enhanced Representation through Knowledge Integration), fuels the Ernie Bot, a chatbot aimed to compete with OpenAI’s ChatGPT. Baidu asserts that ERNIE 4.0 matches GPT-4’s capabilities, demonstrating sophisticated understanding and reasoning abilities.

Alibaba: Alibaba’s AI model, Tongyi Qianwen (commonly known as Qwen), is a comprehensive set of foundational models adept at a range of tasks, from generating content to solving mathematical problems. Select versions of Qwen are open-source, enabling developers to utilize and modify them for various uses. Alibaba has announced that Qwen models are in use by over 90,000 enterprise clients.

Tencent: The Hunyuan model from Tencent is a prominent component of China’s AI landscape. Offered through Tencent’s cloud computing division, Hunyuan is tailored to facilitate a broad spectrum of applications, encompassing natural language processing and computer vision.

Huawei: In spite of considerable obstacles stemming from U.S. sanctions, Huawei persists in AI innovation. The firm has created its own AI processors, like the Kunlun series, to diminish dependence on international technology. Huawei’s AI features are incorporated into a diverse array of products, including smartphones and cloud solutions.

Comparison to the U.S.

The U.S. continues to be a dominant force in AI, with leading companies such as OpenAI, Microsoft, Google, Anthropic and Meta spearheading advancements.

Generative AI: U.S. firms have advanced significantly in generative AI, with OpenAI’s GPT-4 and Google’s Gemini at the forefront. These models excel in creating text, images, and videos from user inputs. Although Chinese models like ERNIE and Qwen are strong contenders, the U.S. maintains a slight lead in capabilities and market penetration.

Semiconductor Design: The U.S. leads the semiconductor design industry, vital for AI progress. U.S. companies command an 85% global market share in chip design, crucial for AI model training and system operation. China’s dependence on imported semiconductors is a notable obstacle, but there are ongoing efforts to create homegrown solutions.

Research and Innovation: Both nations boast strong AI research sectors, yet the U.S. edges out slightly in generating state-of-the-art AI products. U.S. tech giants frequently introduce AI breakthroughs to the market, with Chinese firms quickly gaining ground.

Government Support: The Chinese government ardently backs AI advancement, enacting strategies to spur innovation and lessen foreign tech reliance. Such support has spurred China’s AI industry’s rapid expansion, positioning it as a strong rival to the U.S.

Conclusion

The competition in AI development between China and the U.S. is escalating, as both countries achieve significant breakthroughs. Although the U.S. maintains a marginal lead in some respects, China’s swift advancement and state backing indicate that the disparity might keep closing. The quest for AI dominance by these nations is set to influence the worldwide technological and innovative landscape profoundly.

As of September 2024, it is estimated that China’s AI development is approximately nine months behind that of the U.S.