←back to thread

Z-Image: Powerful and highly efficient image generation model with 6B parameters

(github.com)
396 points doener | 2 comments | 30 Nov 25 11:36 UTC | HN request time: 0s | source
Show context
vunderba ◴[06 Dec 25 17:36 UTC] No.46175068[source]▶
I've done some preliminary testing with Z-Image Turbo in the past week.

Thoughts

- It's fast (~3 seconds on my RTX 4090)

- Surprisingly capable of maintaining image integrity even at high resolutions (1536x1024, sometimes 2048x2048)

- The adherence is impressive for a 6B parameter model

Some tests (2 / 4 passed):

https://imgpb.com/exMoQ

Personally I find it works better as a refiner model downstream of Qwen-Image 20b which has significantly better prompt understanding but has an unnatural "smoothness" to its generated images.

replies(6): >>46175104 #>>46175331 #>>46177028 #>>46177043 #>>46177543 #>>46178707 #
1. rendaw ◴[07 Dec 25 02:45 UTC] No.46178707[source]▶
That's 2/4? The kitkat bars look nothing like kitkat bars for the most part (logo? splits? white cream filling?). The DNA armor is made from normal metal links.
replies(1): >>46178934 #
2. vunderba ◴[07 Dec 25 03:36 UTC] No.46178934[source]▶
Fair. Nobody said it was going to surpass Flux.1 Dev (a 12B parameter model) or Qwen-Image (a 20B parameter model) where prompt adherence is strictly concerned.

It's the reason I'm holding off until the Z-Image Base version is released before adding to the official GenAI model comparisons.

But for a 6B model that can generate an image in under 5 seconds, it punches far above its weight class.

As to the passing images, there is white chocolate kit-kat (I know, blasphemy, right?).