- This topic has 0 replies, 1 voice, and was last updated 4 months, 2 weeks ago by
.
Viewing 0 reply threads
-
Posts
-
-
Hi Andrew,
We’re lucky to have new models appearing all the time and I’ve been trying Qwen-2512 which is a big step up from Qwen 3.0 in terms of image quality and also better on prompt adherence. It would be great to have your expert comparison of it against, say, Z-Image and Flux.2-Dev, both of which run reasonably quickly on A100 in the quantized versions.
A new, faster Flux.2-Klein-Base model has just been released in 4B and 9B versions and I’m trying it out at the moment. Again, good to have a post on that at some point.
Thanks
-
Viewing 0 reply threads
- You must be logged in to reply to this topic.