Alibaba Claims Qwen3.8 Is World’s No.2 AI Model
Alibaba has claimed that its new Qwen3.8 model is the second best Artificial Intelligence (AI) model globally, right after Anthropic’s Fable 5. However, they have not provided any benchmark proof to support this assertion, and the open weights for the model are still weeks away from release.
Key Points:
- Model Details: Qwen3.8 boasts 2.4 trillion parameters, making it the first multimodal model in the Qwen family with capabilities across text, images, video, and documents.
- Performance Claim: Alibaba claims Qwen3.8 outperforms its previous flagship, particularly in coding and office work like full-stack development and data analysis.
- Lack of Evidence: The claim was made without benchmarks, a model card, or independent testing, which is a departure from the company’s usual practice of publishing results for their flagships.
- Open-Weight Strategy: Following a recent trend in Chinese AI labs, Alibaba plans to release Qwen3.8 through paid products first and make the open weights available later.
- Competitive Landscape: The launch comes after Moonshot released Kimi K3, a powerful open model that has sparked market reactions.
In essence, Alibaba’s bold claim about Qwen3.8 represents a significant move in the AI competition landscape, especially within China.