Alibaba Unveils Qwen 3.8-MAX AI Model as DeepSeek Launches Ultra-Low-Cost Alternative

Alibaba has unveiled the Qwen 3.8-MAX AI model, while DeepSeek has released a new AI model described as the cheapest to run among well-known models. Alibaba's model's capabilities are yet to meet the 'second only to Fable 5' claim. This follows China's ongoing challenge to Silicon Valley's AI dominance.

Alibaba unveiled Qwen3.8-Max on August 3, its largest and most capable AI model to date: a 2.4 trillion-parameter mixture-of-experts system that activates roughly 95 billion parameters per query and supports a 1 million-token context window. The model handles text and vision and is aimed at coding, research and long-running autonomous agent tasks, with weights slated for release on Hugging Face and ModelScope in the coming week.

The launch matters because it sharpens the competitive gap between Chinese and US frontier labs. On the Arena.AI leaderboard, Qwen3.8-Max became the highest-ranking Chinese model for text tasks, though it still trails several offerings from Anthropic, including Claude Fable 5. For vision tasks it ranked second globally, behind only a Fable 5 variant. Alibaba's own benchmark disclosures claim scores that match or beat Fable 5 on specific tasks, even as the model falls short of unseating Anthropic's flagship overall.

Separately, DeepSeek released its V4-Flash model on August 1, pricing it at $0.14 per million input tokens and $0.28 per million output tokens, more than 100 times cheaper to run than Fable 5. Artificial Analysis scored V4-Flash at 50 out of 100 on its Intelligence Index, reflecting a real tradeoff between cost and raw capability; DeepSeek is reportedly preparing a more powerful V4-Pro variant. Together, the two releases underscore China's two-track AI strategy: chase frontier capability with Qwen while undercutting Western pricing with DeepSeek.

For BABA investors, the model launch is a near-term sentiment catalyst that reinforces Alibaba Cloud's AI narrative, while the broader competitive dynamic with DeepSeek and Anthropic is worth watching for enterprise adoption trends and further price competition in the AI inference market.