Alibaba’s Qwen takes on Kimi K3 with open-weight Qwen 3.8, says model is “second only to Fable 5”

What happened

Alibaba has launched Qwen 3.8, a new open-weight multimodal AI model with 2.4 trillion parameters. The Qwen team markets it as a top-tier alternative to existing large multimodal models, ranking it just behind Fable 5 and positioning it as a competitor to models like Kimi K3. A preview of Qwen 3.8 is available now, allowing early access to developers and researchers.

Why it matters

Qwen 3.8’s scale and open-weight release signal Alibaba’s clear push to challenge dominant AI players with a model designed for both text and image inputs. Its size and multimodal capability suggest strong performance potential, which raises the stakes in Asia’s expanding AI market and globally. An open-weight approach lowers barriers for enterprises and developers looking to customize or build on a sophisticated model without the constraints of closed proprietary systems. Having a viable open competitor can pressure incumbents on licensing, pricing, and innovation pace.

What to watch next

Keep an eye on how Qwen 3.8 performs when tested beyond Alibaba’s environment, particularly for real-world applications in automated customer service, content generation, or multimodal workflows. Observe whether the open-weight strategy gains traction among startups and enterprises needing adaptable, large-scale AI without vendor lock-in. Watch competitors’ responses, especially updates to Fable, Kimi K3, and other Chinese or global models aiming to secure dominance in multimodal AI offerings and developer ecosystems.

AI Quick Briefs Editorial Desk