Alibaba has introduced its latest AI model, Qwen3.5, which boasts superior performance at a lower cost compared to its predecessor. The model is designed with "visual agentic capabilities", enabling actions across various apps on phones and computers. With a focus on optimizing speed without compromising capability, Qwen3.5 has 397 billion parameters, but only activates 17 billion per forward pass.

In a bid to compete with US counterparts in the AI domain, Alibaba's Qwen3.5 outperforms OpenAI's GPT-5.2, Anthropic's Claude Opus 4.5, and Google's Gemini 3 Pro in various benchmarks. This latest launch underscores the escalating competition between Chinese and American tech giants striving for AI superiority.

The unveiling of Qwen3.5 follows similar moves by other Chinese tech firms. ByteDance recently upgraded its Doubao chatbot app and introduced Seedance 2.0, an AI video generator. Additionally, Zhipu launched GLM-5, MiniMax released M2.5, and Moonshot AI presented Kimi K2.5. These developments precede the upcoming release of DeepSeek's V4 model, anticipated to outperform competitors ChatGPT and Claude in tasks involving lengthy coding prompts.

As China solidifies its position as a key player in the global AI landscape, the introduction of Qwen3.5 by Alibaba signifies a significant stride in the country's pursuit of AI advancement. The rapid pace of innovation in Chinese tech companies underscores the fierce competition and continuous quest for technological supremacy in the realm of artificial intelligence.