News

Chinese tech Giant Alibaba Releases Coding AI Model Amid US-China Tech Rivalry

• By John K. Waters
• 07/24/2025

Alibaba Group launched an open-source artificial intelligence model for software development on Wednesday, marking the latest move by Chinese technology companies to compete with U.S. rivals in the rapidly evolving AI sector.

The e-commerce giant's Qwen3-Coder model is designed to handle complex programming tasks, including code generation and debugging entire software systems. The release comes as Chinese firms face mounting pressure from U.S. export restrictions on advanced semiconductors and AI technology.

Alibaba said the model, which contains 480 billion parameters but activates only 35 billion at a time for efficiency, outperformed domestic competitors, including DeepSeek and Moonshot AI, in coding benchmarks. The company also claimed it matched the performance of leading U.S. models from Anthropic and OpenAI in certain areas.

The model's open-source nature reflects a strategic shift among Chinese AI companies, which are increasingly releasing their technology freely to build developer communities and compete with proprietary U.S. systems. This approach contrasts with the closed development model favored by many American AI firms.

"Agentic AI coding is transforming software development by enabling more autonomous, efficient, and accessible programming workflows," Alibaba said in a statement.

The launch highlights the intensifying competition in AI development between Chinese and U.S. technology companies. Although American firms like OpenAI and Anthropic have dominated headlines with their chatbot models, Chinese companies have focused heavily on specialized applications, such as coding and scientific research.

Qwen3-Coder can process up to 256,000 tokens of context—roughly equivalent to analyzing several hundred pages of code simultaneously—and can be extended to handle up to 1 million tokens. This capability allows developers to work with large, complex software projects within a single session.

The model will be available through Alibaba's cloud platform and popular developer repositories including GitHub and Hugging Face. Alibaba's existing coding assistant, Tongyi Lingma, has generated more than 3 billion lines of code since its launch in June 2024, according to the company.

The release follows a pattern in which Chinese AI companies make their models freely available, while U.S. competitors often charge for access to their most advanced systems. This strategy aims to build market share and gather user feedback, though it raises questions about long-term monetization.

Chinese technology companies face ongoing challenges from U.S. export controls on advanced semiconductors, which are crucial for training and running large AI models. These restrictions have prompted Chinese firms to focus on more efficient model architectures and seek alternative chip suppliers.

The global AI market is expected to reach $1.3 trillion by 2032, according to research firm BloombergNEF, with coding applications representing a significant segment as companies seek to automate software development tasks.