Raehoshi Character List
Author: Raelina
Model Link: Raehoshi illust XL
Changelog:
[2025.12.10] Initial release. Character list supports up to v8 and Vpred v1.
[2026.02.18] Updated Character list for v9 and Notes.
Introduction
The Raehoshi Character List is now officially available.
This document includes every character I personally trained based on the dataset used during model development.
All character names listed here correspond to the versions available up to v9. For details on changes or additions in each version, please review the version notes provided on the respective model pages.
This list is intended to support users in achieving more accurate prompt construction and improving generation results when working with Raehoshi models.
Important Notes and Disclaimer
Character Scope: This list contains only characters trained specifically by me. Characters natively supported by the Illustrious or NoobAI base models are not included here.
Generation Complexity: Some characters may be difficult to generate, particularly in their default outfits. This is often due to limited available data or a lack of diverse fanart for those specific forms during the training process.
Prompting Requirements: This list provides character names only. To achieve the best results, you must manually specify key attributes such as:
Hair color and hairstyle
Eye color
Specific accessories or outfit details
Usage and Attribution:
Model Merges: If you include this model in a merge, please provide clear credit to the original work.
Reference Lists: If you use this character list on your own pages or documentation, please link back to this article as the source.
Support the Project
If you would like to support ongoing Raehoshi development, please consider contributing via Ko-fi.
Your support directly helps cover cloud GPU rental costs and ensures continued updates and improvements to the project.
Additional Resources
Wildcard files are included in the attachment.
Categorized character lists by series are available here: rentry
This list is based solely on the training dataset.
Not all characters have been individually tested yet, so feedback regarding difficult or inconsistent outputs is welcome and appreciated.