Alibaba's Qwen Launches Qwen3.8-Flash with 89% Lower Training Cost
BABA sits 29% above its 52-week low of $91.99.
Summary
Alibaba's Qwen lab released Qwen3.8-Flash, a multimodal model that cuts training costs to about one-ninth of the prior Qwen3.7-Plus while improving coding and office-task performance. The model supports a 262,144-token context window expandable to 1 million tokens, and API pricing is set at 1 yuan per million input tokens and 3 yuan per million output tokens. Qwen also open-sourced weights for Qwen3.8-Flash-Next, a prototype for the upcoming Qwen4 family. This launch follows Alibaba's record HK$80 billion share placement on Sunday to fund AI infrastructure, and comes amid intensifying AI competition in China. The cost reduction and open-source move strengthen Qwen's competitive position, but the immediate financial impact is modest relative to Alibaba's scale.
At the time of this announcement, BABA was trading at $118.94 on NYSE in the Technology sector, with a market capitalization of approximately $279.3B. The 52-week trading range was $91.99 to $192.67. This news item was assessed with positive market sentiment and an importance score of 7 out of 10. Source: Reuters.