(last edit : 2026/06/26)
This article covers LoRA models I plan to create or accept as requests.
Please feel free to make requests within reasonable bounds. I would greatly appreciate it.
✅ : approve 🤔 : low priority ❌ : reject
📢Important Notes
I am mainly interested in VTubers, so please understand that I am more likely to prioritize female VTuber LoRAs due to personal preference.
👥If a LoRA that reproduces the target already exists with the same model, I will not create it.
Please kindly check whether it already exists on Civitai before making a request.Likewise, I kindly ask other LoRA creators not to create duplicates of the LoRAs which have made by me or others.
This should be considered part of the ethics of LoRA creation.As an exception, if an existing LoRA does not cover all outfits, I may create one as long as it doesn't overlap with the existing version.
(Example: For Yuuki Sakuna, there is already an Illustrious LoRA covering the first outfit, but not the second, so I can create one for the second outfit.)Exceptionally, if existing LoRAs only cover a single default outfit subject, the new LoRA I'll make can cover default outfit also.
🔄If a target can already be sufficiently reproduced by the base model, it will be given very low priority.
As another exception, if certain outfits cannot be reproduced properly by the base model, I can still create a LoRA.
(Example: Yakumo Beni — her appearance can be reproduced by the base model, but her outfit cannot, so I created a LoRA.)
❌Requests may be rejected in the following cases:
If there are fewer than 10 sufficiently high-quality images that are not blocked by AI training restrictions
If the target to be reproduced is male or non-human characters
If the target is explicitly marked as not allowed for AI training
If the target has too many outfits (over 12+ outfits)
Instead, in this case, I can accept if you specify a specific outfit.
🧩For base models:
✅By default, I mainly focus on creating Illustrious version LoRAs.
❌As for SD 1.5 and Pony versions, I am currently not accepting requests for them, as there is very little demand for them at this time.
It may take 1–2 months or longer to complete a request. Please keep this in mind.
The order in which requests are submitted does not guarantee the order in which they will be completed.
This depends on a combination of my personal preferences and the difficulty of collecting datasets.
Normally, LoRA models are planned to be produced once every 2–3 days, but this schedule may change due to personal circumstances.
My preferences are as follows (items listed earlier are preferred more):
Human > Halo or Wings > Animal ears > Tail > Horns > Pointed ears
The vtuber’s canonical race is not important. Even if a vtuber claims to be an alien, they will be treated according to human-based preferences as long as they look human.
More recently debuted vtubers, and those with fewer outfits, tend to be completed faster.
Therefore, vtubers who debuted before 2021 and have many outfits (4 or more) will likely be completed very late.
For Twitch streamers, if there is little fan art, the likelihood of being made is extremely low. I cannot watch Twitch VODs.
If the backlog of requests becomes too large, accepting new requests may be temporarily suspended.
If you make a large amount of LoRA requests at once, my backlog will be jammed. Therefore, I will limit the maximum number of items that can be requested to 3 items at once.
(When your request is completed, it doesn't matter to make a new LoRA request.)
📝Plans
This section covers the LoRA models I plan to create or update (including requests).
🎨 : illustrious 🦄 : pony 🐥 : noobai 👽 : other models
🚀 : initiative 🛎️ : request
✨new LoRAs
🎨🛎️Shirakami Fubuki (after solo live outfit)
🎨🛎️Ookami Mio (after gamer outfit)
🎨🚀Vestia Zeta (all outfits)
🎨🛎️yono
🎨🛎️mikage
🎨🛎️lewne
🎨🛎️meda
🎨🛎️imi
planned, but low priority
I have a plan to produce these to the end, but it is likely to be completed very slowly
🎨🚀Amahane Yotsuha (all outfits)
🎨🚀nijisanji idol outfit
🎨🚀vspo! showdown(2025) outfit
🎨🚀Sakusan Canon
🎨🚀Washu Asuka
🎨🚀Momohina Nano
🔧update LoRAs
I have a plan to produce these to the end, but it is likely to be completed very slowly
🎨🚀Suzutsuki Sui - 2.0
🎨🚀Yakumo Beni - 2.0 (rekt, maybe will re-train this)


