Chinese Tech Giant Accelerates Global AI Race with New Model Release

Alibaba Cloud has launched its new open artificial intelligence model, Qwen-3.5, marking a significant impact in the global AI race between China and the US. This highly anticipated release coincided with the eve of the Lunar New Year, amidst a week where numerous leading Chinese AI developers introduced their flagship models.

Details of the Qwen-3.5 Models

The two latest models, available on Alibaba’s cloud platform Model Studio, are Qwen-3.5-Open-Source and Qwen-3.5-Plus. The open-source model features 397 billion parameters, reflecting notable advancements over its predecessor, Qwen-3-Max-Thinking, which has over 1 trillion parameters. Self-reported benchmark scores indicate that Qwen-3.5-Open-Source delivers significant improvements in performance.

Performance Comparisons

According to the benchmarks, Qwen-3.5’s performance is comparable to leading models from OpenAI, Anthropic, and Google DeepMind. However, it is worth noting that these comparisons do not involve the most recent versions of the US competitors’ models.

Features of Qwen-3.5-Plus

The Qwen-3.5-Plus model, which operates under closed-source conditions, is reported to have performance levels that align with top-tier models in the industry. It boasts a context window of 1 million tokens, providing one of the largest data processing capacities available.

Implications for the Global AI Landscape

The launch of Qwen-3.5 models by Alibaba Cloud emphasizes the growing competitive landscape in AI technology. With significant advancements made by Chinese tech firms, this release accelerates the ongoing race in artificial intelligence development globally.