Model has a stronger anime style and more consistent backgrounds than base Pony Diffusion V6 model making it easier to prompt. It is compatible with Lora trained on Pony Diffusion and includes the same characters and styles in addition to ones added by Confetti.

Characters included in the pony base model have their own style which you can try to override with artist prompts or style loras.

Recommended positive prompt:

score_9, score_8_up, score_7_up, source_anime

Optionally add rating_safe ,rating_questionable or rating_explicit

Optionally add score_6_up, score_5_up, score_4_up

Including these last three score tags will weaken your lora's style. You can also try removing source_anime if your lora has a strong style.

Optional negative prompt:

ugly, sketch, 3d

Look at the x/y plot images in the gallery (https://civitai.com/images/7559185) to see examples of how positive and negative prompts affect the image. The x-axis is the positive prompt and the y-axis is the negative prompt.

This x/y plot (https://civitai.com/images/8112483) compares different negative prompts so you can see their effect on the model

Style can be applied by using art by (name)

Look at attached images for examples of how this modifies the image.

Each version merges a newer version of Anime Confetti Tune but doesn't necessarily produce better results.

This x/y plot (https://civitai.com/images/8112489) compares the two base models and the three versions.

animeconfettitunexl_con5tixl and comradeshipXL_v5aDPO are the base models that were merged.

anim3conf3tticomrade is v1

confetti4comrades is v2

confetti5comrade is v3

All images are generated with Euler A sampling method, 6 CFG Scale using WebUI Forge.

WebUI Forge is built on top of Automatic1111's backend but has lower VRAM usage and higher performance. Try it out!