Updated: This lora is considered unsuccessful due to the different subtle styles mixing caused by my inexperienced in choosing images when I first getting into ai image.

The one of the reasons why I started my ai image generation journey. Truck kun’s 2025 Apr-June style is just way too good. Other than obvious mix of Hews in his style, everything else is completely unfindable. So, I took his images make this lora specifically trained using that one specific style among other variant in his 2025 June image set. Might generate weird body proportion or not adhering to prompt.

Trained on Pixai using 100 images.

Weight: 1.0-1.3. Either Euler or Euler a is fine, but preferably Euler.
Style strongly depends on CFG scale
Tested CFG:
- 3.0-4.0 for that exact heavy outlines style
- 2.0-2.5 for pretty much subtle enough for other style.

See cover images for differences, though generated using VXP_ASC vPred (Anime ScreenCap), but you see that can this lora can replicate the style with the right checkpoints.

Tested using following models: (low steps & resolution because it's cost less token to test result)
First set of examples use:
- VXP_ASC vPred (Anime ScreenCap)
Second set of exanples use:
- VXP IL Base Model (NSFW alert!)

(maybe) - WAI v17

Or

any other better checkpoints or control settings of your preferences.

Enjoy!