V0.5: This version is mainly a anime style with soft shading, slight details and a extremely slight semi-realistic aesthetic. It also can feature a slight gritty feel depending on the image you receive. Feel free to post any image you want, long as it's ToS friendly (SFW/NSFW).
V0.7: This version is mainly a semi-realistic style with soft shading, slightly bad adherence, very slight soft saturation with a softer style. It also has a bit more of a dark fantasy feel. It also might struggle with emotions slightly. Feel free to post any image you want, long as it's ToS friendly (SFW/NSFW). Needs Pony Scores on simple prompts.
V1.0: This version is mainly a semi-realistic style with better adherence that combines some of the soft shading from V0.7 with slight film grain (like 1% or something), with a little bit of dark fantasy aesthetic. Feel free to post any image you want, long as it's ToS friendly (SFW/NSFW).
- Descriptions by mostly ChatGPT.
This model is only supported on CivitAI and Tensor.Art. Also all my models I made for personal use but I upload them to give back to the community.
Recommended for V1.0:
Sampler: Euler
Scheduler: Normal
CFG: 6
Steps: 20 - 30 Steps
Note: You may need to regenerate a couple of times or do batches as sometimes it's not 100% accurate.
Recommended for V0.7:
Sampler: Euler A
Scheduler: Normal
CFG: 4 - 6
Steps: 20 - 30 Steps
Note: You may need to regenerate a couple of times or do batches as sometimes it's not 100% accurate.
Recommended for V0.5:
Sampler: Euler A
Scheduler: Normal
CFG: 4 - 6
Steps: 20 - 30 Steps
Note: You may need to regenerate a couple of times or do batches as sometimes it's not 100% accurate.
Merges Using This Merge Model (Please let me know if you make a merge with this so it's added to the list.):
⚠️ IF YOU'RE GOING TO STEAL MY MODELS, MAKE IT SO THE DOWNLOAD IS ENABLED. THIS WORKS BETTER LOCALLY. ⚠️
Description
This version focuses on a more semi-realistic 2.5D style with a softer aesthetic, with also a slight dark fantasy aesthetic. Slight incorrect prompting may lead to the loss of certain tokens.
# New Stuff Merged In:
LoRA specifically trained for v0.7 and on V0.5.
A second LoRA specifically trained for v0.7.
A third LoRA specifically trained for v0.7.
Detail Slider LoRA | PonyXL SDXL - V1.4 (LoRA)
Note: Block Merged all that into this version.
Spent 2480 buzz for all three LoRAs.
FAQ
Details
Files
Available On (2 platforms)
Same model published on other platforms. May have additional downloads or version variants.