Recommended Prompts
(masterpiece, best quality:1.4), (beautiful, perfect, delicate, detailed, intricate, aesthetic:1.2), ...
Recommended Negative Prompts
easynegative, badhands, badartist, badartistanime, badprompt, (low quality, worst quality:1.4), (monochrome, grayscale, sketch:1.2), (logo, signature, text, words, username, artist name, watermark), (loli:1.2), ...
Recommended Parameters
samplers | DPM++ SDE Karras | |
steps | 30 ~ 50 | |
cfg | 7 ~ 8 | |
clip skip | 2 | |
vae | kl-f8-anime2 VAE | |
resolution | 512×768 to 1024×1536 |
Recommended Hires (High-Resolution) Parameters
upscaler | 4x-UltraSharp | |
upscale | 2 | |
steps | 6 ~ 10 | |
denoising strength | 0.5 |
Tips
- When creating images, use the stylized Delta version for colorful and detailed outputs.
- Use the Swamp version for realistic proportions combined with vibrant details.
- For the most realistic facial expressions and more intricate details, opt for the Soul version.
Version Highlights
Soul version with baked VAE
Creator Sponsors
All sponsors are not affiliates of Diffus. Diffus provides an alternative online Stable Diffusion WebUI experience.
The Model MixTape Blus is similar to the MixTape Rock n’ Roll, using the model DuckHaitenNiji in its base, but with a different mix. It’s not very good with your hands, but it creates images with rich details and colors.
The Delta version is the most stylized. It creates very colorful images, with lots of detail and a more three-dimensional finish.
The Swamp version has more realistic proportions but still creates stylized, colorful and very detailed images. Features more facial expressions and more varied backgrounds.
The Soul version creates people with the most realistic proportions. Makes good facial expressions, lots of details. Create some very naughty scenes 😉
Recommends
Positive Prompt:
(masterpiece, best quality:1.4), (beautiful, perfect, delicate, detailed, intricate, aesthetic:1.2), ...
Negative Prompt:
EasyNegative, BadHands, BadArtist, BadArtistAnime, BadPrompt, (low quality, worst quality:1.4), (monochrome, grayscale, sketch:1.2), (logo, signature, text, words, username, artist name, watermark), (loli:1.2), ...
Embeddings (Textual Inversion): EasyNegative, BadHands, BadArtist, BadArtistAnime, BadPrompt
Sampling method: DPM++ SDE Karras
Sampling steps: 30 ~ 50
Hires. fix:
Sampler: 4x-UltraSharp
Hires steps: 6 ~ 10
Denoising strength: 0.5
Upscale by: 2 (resize from 512×768 to 1024×1536)
Clip skip: 2
CFG Scale: 7 ~ 8
VAE: kl-f8-anime2 VAE
Merge Recipe
Merged models are the following:
Delta Steps:
-
Base = GoldenHarvest x 0.7 + FantasticChixHR x 0.7
-
Delta = Base x 0.3 + DucHaitenNiji x 0.7
Swamp Steps:
-
Base = GoldenHarvest x 0.7 + FantasticChixHR x 0.7
-
Delta = Base x 0.5 + DucHaitenNiji x 0.5
Soul Steps:
-
Base = GoldenHarvest x 0.7 + FantasticChixHR x 0.7
-
Delta = Base x 0.7 + DucHaitenNiji x 0.3
License
All used models here have “Have different permissions when sharing merges” license permission, so I do not try to add any restriction on it.
This model permits users to:
✔Use the model without crediting the creator
✔Sell images they generate
✔Run on services that generate images for money
✔Share merges using this model
✔Sell this model or merges using this model
✔Have different permissions when sharing merges