🍩 Chubby | LoRA — Base: ANIMA

🙏 Before anything else: this LoRA took a lot of work — picking the images, building the dataset and tags, training, and testing the checkpoints. If you have a moment, drop a Like and especially post your results in the comments (with the prompt/settings you used) — it helps me a lot to know what to improve in the next versions. 🔥

💬 Trained with a balanced mix of and SFW images, all featuring unique, varied, adult-only characters.

💬 This LoRA can be used for generation. Use it respectfully.

💬 Check out how the prompt examples were set up in the posted images.

📦 Releasing two checkpoints — 200 steps and 1600 steps — so you can pick the strength of the effect you want.

📊 About this model

This is not a character LoRA — it's a body-type/ LoRA focused on chubby . The vanilla ANIMA base already tends to generate this kind of body by default, but I built this LoRA around my own preferences for how that should look and behave.

Both checkpoints generate generic characters noticeably better than the plain ANIMA base when it comes to this type of . This model tends to generate strong/muscular characters without the "carved/etched" muscle effect the ANIMA base usually produces, and it also tones down the "plastic" look the base model tends to give this body type. Both checkpoints are solid for generic/general-purpose generation, though I personally prefer the 1600 checkpoint.

⚠️ Highly recommended: use some kind of LoRA (like this one) or well-crafted prompts whenever generating this body type. Without them, the ANIMA base tends to generate noticeably more childish-looking characters than intended.

ℹ️ Generation tips

• On ANIMA base with 12 steps + acceleration LoRA, the output has a hand-drawn animation feel that works great as a starting point

• Using more steps or a different ANIMA-based model with quality LoRAs will improve sharpness and detail

• Higher CFG can help lock in the more consistently

• Check the prompt examples used in the posted preview images — they show how to structure tags for this body type

🤝 Using it together with character LoRAs

Test both weights when combining this with a character LoRA:

• 200 tends to push the image more strongly, shifting it toward a chubbier — though still defined — look

• 1600 is more conservative and preserves the character LoRA's own style more faithfully

If you like the character LoRA's default and just want a light touch, go with 200. If you want a stronger, more defined shift, go with 1600.

👥 Generating with two characters

Same idea — test both weights and see what fits your scene. 200 gives a lighter nudge if you're happy with ANIMA's default proportions; 1600 gives a stronger, more consistent overhaul across both characters.

📊 Checkpoints

Steps

200 Leans toward a chubbier, still-defined look — pushes the image more strongly, more noticeable style shift

⭐ 1600 Leans toward a stronger, more defined/muscular look — my personal favorite

This is the V1 release and is still subject to some errors, mainly around camera angle and framing — it tends to favor tighter/closer shots. I have checkpoints available up to 2000 steps, though there are no real gains beyond 1600. If you'd like something in between 200 and 1600, leave a comment and I can upload it on request.

⚙️ Generation settings used for preview images

All preview images were generated using:

• Base model: Anima 1.0

• Acceleration LoRA: strength 1.0

• CLIP: anima default

• Steps: 12

• CFG: 1.0

• Sampler: er_sde - beta

• Image size: 816x816 (strength-comparison images: 816x1280)

🔮 What's next

I'm planning to post more LoRAs soon — simple poses, poses, and focused more specifically on content. I'm still studying the best way to approach that on ANIMA.

I make these LoRAs based on personal taste, so if this one isn't for you, that's completely fine — I'm just experimenting. If you want something specific, send me a message.

👍 One more thing

I'm just getting started with this — if you have a moment, drop a Like and share your generations with the settings you used. It really helps me improve the next versions. Have fun! 🔥