Qwen 3.8 Max IS INSANE! Second To Fable? New Open Model King? (Fully Tested)

· Source: WorldofAI · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Software Development & Engineering · Depth: Expert, long

Summary

Alibaba has launched Qwen 3.8 Max, a new 2.4 trillion parameter multimodal AI model, with an open-weight release anticipated soon. Initially available as a preview for token plan users and Qwen chat ambassadors, Alibaba claims it rivals leading Frontier models and is second only to Fable 5. However, early assessments suggest it lags behind Fable 5 and Kimi K3, Kimi GPT 5.1.6 Soul, and GPT 5.6 6 Luna. Qwen 3.8 Max offers significant multimodal upgrades, supporting text, images, and videos with a 1 million token context window, and demonstrates strong performance in language and vision. It shows notable improvements in coding, advanced reasoning, and front-end design, generating responsive landing pages and complex SVG code for animated scenes and even a functional MacOS clone. While faster than Kimi K3 in some tasks, it excels in specific areas like realistic deformation physics rendering and open-world game generation.

Key takeaway

For AI Engineers evaluating new open-weight models for creative code generation, Qwen 3.8 Max offers compelling capabilities in front-end development, 3D, and SVG. While it may not outperform Kimi K3 in all benchmarks, its speed and ability to generate complex animated scenes and functional OS clones in a single shot make it a strong contender for specific application development. Consider testing its multimodal features for projects requiring detailed visual rendering or intricate code structures, especially if faster, decent outputs are prioritized over absolute top-tier performance.

Key insights

Qwen 3.8 Max is a powerful multimodal model excelling in complex code generation and visual rendering, despite not matching top-tier benchmarks.

Principles

Method

Qwen 3.8 Max employs a "thinking enabled" mode, utilizing multiple internal loops to refine outputs for complex tasks like deformation physics rendering, leading to exceptional quality but longer generation times.

In practice

Topics

Best for: NLP Engineer, Computer Vision Engineer, AI Scientist, Machine Learning Engineer, AI Engineer

Related on AIssential

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by WorldofAI.