Browse categories
Explore
Fiverr Pro
English
$
USD
I will train an illustrious or anima style lora
Kurut L.
Level 1
About this gig
I will train a custom Stable Diffusion LoRA model for a style of your choice, based on either the ANIMA or Illustrious base model.
I can do models for these categories:
- Anime
- Manhwa/Manga
- Cartoon
- Doujin
- Video game
- Original Creator Style
- ... and more. Message me for special cases!
You tell me exactly which style you are looking for, I handle dataset selection, tagging, training, and testing. If source material is not publicly available, I will need you to provide the material.
All packages include:
- A ready-to-use LoRA file
- Recommended trigger words
- Basic usage guidance
- at least 20 Image previews of the LoRA
- at least 20 Image comparisons (LoRA - No LoRA)
Package overview:
- Basic: 1 style LoRA
- Standard: 2 style LoRAs
- Premium: 3 style LoRAs
The LoRA is specifically optimized for the selected base model (Anima or Illustrious) and is not intended to be base-agnostic. This ensures stronger style accuracy and more predictable results.
Optional add-ons include faster delivery, access to all epoch files, or a private model.
Original styles are possible. If you are uncertain if I would be able to train a LoRA for what you have in mind, please message me before ordering.
Delivery style preference
Please inform the freelancer of any preferences or concerns regarding the use of AI tools in the completion and/or delivery of your order.
Get to know Kurut L.
Kurut L.
LoRA Trainer
Level 1
- FromGermany
- Member sinceJan 2026
- Avg. response time6 hours
- Last delivery1 day
Languages
English, German, Ukrainian, Russian
I specialize in training custom Stable Diffusion LoRA models with a strong focus on character accuracy, consistency, and usability. I handle dataset curation, tagging, training, and testing to ensure clean and reliable results. I mainly work with anime-style characters and understand both SFW and NSFW workflows. You’ll receive a ready-to-use LoRA, trigger words, and usage guidance so you can start generating immediately.
I have released over 2000 models on CivitAI (Username: KurutAI) in the past few months, they are all available for free directly on CivitAI!
FAQ
What do you need from me to start?
The source, description of what exactly of that style is important to you, and whether you want an Illustrious or Anima base. Providing images of that style helps, but isn't required for publicly available style sources.
Can you create a LoRA for any style?
Yes. My training workflow is optimized to capture the core style design, proportions, and visual identity. In rare cases where reference material is extremely limited, I’ll let you know before proceeding.
Are there any limitations I should be aware of?
A style LoRA will only turn out perfect, if there's enough different source material. If you only have images of a single character in that style for example, the style LoRA will have a hard time generating other characters.
Which base model do you use for training?
All LoRAs are trained on either Anima or Illustrious. They will work on all Anima or Illustrious Checkpoints, results may vary depending on the checkpoint.
How flexible is the LoRA after training?
You have a very limited amount of space for adjustments. While you can reduce the weight of the style LoRA, you will also reduce the effectiveness of the style lora, which will result in a less accurate style.
What are epoch files?
Intermediate versions saved during training, useful for testing or personal fine-tuning.
Can I commission a multi-style mixed LoRA model?
No. I only create single-style LoRA models. Combining separate style LoRAs in one image typically produces better results and higher flexibility than multi-style training.
Can I use the LoRA together with other character or style LoRAs?
Yes. The model is intended to work alongside other LoRAs. As with all LoRA combinations, weight balancing and prompt structure will affect results, especially if you combine multiple style LoRAs.
I have another question
Message me! I'm always happy to answer any questions you might have or discuss LoRA possibilities. I won't judge either.
