Qwen 4 Kalmia 3D test: still behind DeepSeek V4.1 Flash on the same task
After 10+ 3D tests the author finds Qwen 4 Kalmia decent but still below DeepSeek V4.1 Flash on the same task, roughly MiniMax 3.1 Flash level overall.
Qwen 4 Kalmia on Arena, this is the checkpoint from yesterday and I have to say it's pretty good at 3D, although I don't think it's quite at the level of the best models yet, DeepSeek V4.1 Flash still does a better job in this particular test, but we also have to keep in mind that we don't know what reasoning level Qwen 4 was using here, which could make a big difference. After running more than 10 tests with this model, my first impression is that it'll probably be around the level of MiniMax 3.1 Flash, which isn't bad at all! I'm hoping they keep optimizing it and improving its 3D capabilities before the official release, because there's definitely a lot of potential here. And with even more models expected to launch this week and next, we have some exciting days ahead of us :)