春秋@SOG
8.9K posts

春秋@SOG
@immolational
洞察组——The situation observation group chinese

JUST IN: 🇺🇸🇨🇳 Trump administration weighs strict curbs on Chinese AI models after Moonshot Kimi K3 launch, including U.S. host liability and backlist threats.

> Deepseek is creating a MONSTER > they're actively distilling Fable 5 right now > and even involving their own users in this > those who sent requests to Deepseek yesterday sometimes received responses from Fable 5 > the next Deepseek model will probably be very powerful






asked fable to create a virtual PC in a bedroom in @threejs where i can play cyberpunk 2077, this is what it made. pretty cool!



Qwen3.8 is launching and going open-weight soon!🌐 With a massive 2.4T parameters, this model is continuously evolving. We believe it’s one of the most powerful model available today, compatible to leading frontier AI models , second only to Fable 5. You don't have to wait to test it. Just now, the Qwen3.8-Max-Preview made its debut on Alibaba’s Token Plan, Qoder, and QoderWork. Be among the very first to try it out. Can't wait to hear what you build. Stay tuned! 🚀 Token Plan international:qwencloud.com/pricing/token-… China:platform.qianwenai.com/pricing/token-…

Qwen 3.8 Max DESTROYS GPT-5.6 Sol? Polymarket gives Qwen a 70% chance of being the best Chinese AI in July Qwen3.8-Max is now available for use and the initial results are disappointing on the frontend, it performs worse than Kimi K3, Fable 5, and GPT-5.6 Sol in coding tasks, it shows results at best on par with Opus 4.8/GPT-5.5 and yet, Alibaba still claims that they are second only to Fable 5 so far, it doesn't look like Kimi K3 v2



One Shot GTA 6 Qwen3.8-Max-Preview, only generated by the model, no external assets. The best one shot I've seen so far, Qwen is pretty strong in GTAs.











