🧶 Krea 2 Felt Figurines turns one subject into a cute needle-felted wool figurine: fuzzy fibers, soft matte wool, small stitch lines, plump pastel shapes, shot like a handmade toy on a plain backdrop.

✨ Trigger word: feltworld (put it at the start of the prompt)
🧩 Base model: Krea 2 (tested on Krea 2 Turbo; trained on Krea 2 Raw)

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

🎯 What it is good at: ONE object on a plain backdrop
Animals, birds, fruit and sweets, plants, everyday objects, a small vehicle or tent. Prompt pattern that worked (very short, no style paragraph):
feltworld, <one subject>

🖼️ Example prompts (all from the gallery, unedited)
• feltworld, a plump brown owl
• feltworld, a gray koala
• feltworld, a brown snail
• feltworld, a blue bicycle
• feltworld, a slice of watermelon

⚙️ Settings that worked for me (Turbo)
• LoRA strength 1.0
• 10 steps, CFG 1, sampler euler, scheduler simple
• Size tested: 1088 x 1920 (portrait)
• I did not test other strengths, other sizes or the Raw model with this file.

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

📊 What my tests showed (honest numbers)
• 10 short prompts (4 subjects from my held-out set + 6 new ones), written and locked before training. This file (training step 768): 8 of 10 passed my rule (needle-felted wool look, one object, plain backdrop, no face on a non-animal, no extra props). 1 was borderline (a donut with a soft window panel behind it) and 1 failed (a wagon that came with a small figure sitting in it).
• The same words without the trigger gave 0 of 10 felt results (all real photographs). The base model knows the word "felt", but without the trigger it gives a photo scene. The value of this LoRA is short prompt + trigger = one felt figurine on a plain backdrop. It does not "teach felt".
• The final training step (960) passed only 3 of 10 (faces appeared on a donut and a watermelon, outdoor scenes on a snail, a tent and a bicycle), so I publish step 768.
• 10 prompts is a small sample, and I did not check neighbouring steps or other seeds. Expect most, not all, of your results to be clean.

⚠️ Known limits
• One object only. Wide scenes (streets, rooms, landscapes) are not claimed and were not tested.
• Props and companions can sneak in. A "wagon" prompt gave a little figure with a face sitting in it. Try another seed or name just the object.
• Faces on food or objects can appear on some seeds. Add "no face" or try another seed.
• Small training set, one look: it is a style, not a guarantee for every subject.
• No people were used in training or in my tests.

🛠️ About the training
• 20 synthetic images (animals, birds, food, plants, everyday objects, vehicles and small buildings) generated with Krea 2 by me, each with a written caption; 4 more were held out for testing. No scraped, third-party or real-person images.
• LoRA rank 32 on Krea 2 Raw, trained locally with musubi-tuner. File: checkpoint at step 768 of a 960-step run.

🔗 More from me (same base model, Krea 2)
• Krea 2 Clay Figurines: the same idea in modeling clay (https://civarchive.com/models/3022565)
• Krea 2 Origami World: the same idea in folded paper (https://civarchive.com/models/3019950)

📝 Notes
All gallery images are AI-generated with this LoRA (Krea 2 Turbo, strength 1.0), not edited, with their prompts and seeds in the image metadata.

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

⭐ Found this useful?
• Like if it saved you time
• Comment your results, I read every one
• Follow for new ComfyUI workflows, all tested on 16 GB VRAM

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

📜 Licence and notice
Krea 2 is licensed under the Krea 2 Community License Agreement. For more information, visit https://krea.ai/krea-2-licensing.
This LoRA is a modified version (a Derivative) of the Krea 2 model. It is not an official Krea product and is not endorsed by Krea. Use of this LoRA and of the images it makes is subject to that Agreement and to the Krea Acceptable Use Policy (https://krea.ai/krea-2-use-policy). Commercial use is only allowed under the Agreement's company-wide revenue limit of US$1,000,000 per year (section 2.3).