âī¸Model Hub Beta
Last updated
Last updated
You may now upload your own models & spells to use with the image generator.
Visit Model Hub: https://moescape.ai/models/
Models - called Checkpoints
Spells
LoRA
Textual Inversion - mainly used for negative prompts embedding
LyCORIS - basically a multi-layered version of LoRA that is a more advanced and better-trained form
Note:
When uploading SDXL models or spells, please make sure to select 'SDXL' in the 'Category' section. (see 2nd screenshot)
We currently only accept .safetensors files.
If an existing uploaded model is yours, you will be able to put in a claim request directly from the specific model's page, which will then be reviewed by our team.
Go to Model Hub page > click '+ Create Model' button.
Fill out and select the relevant information.
Model Name, Type, Category, Hashtags, etc.
Add version information
Note: Please make sure to select the correct model type as explained above.
Upload model file.
Upload relevant photos for your model. (Your model upload will fail if it contains any explicit images. You may attach multiple, but the first image in the album will be used as the cover image.
Once your upload has been processed, you will be able to use it in the image generator.
Note:
Processing time varies based on file size.
Spell files are typically smaller and can be available almost instantly.
Model files are larger and therefore require more time to process.
The first few generations with newly uploaded models may take longer than usual, as it takes time for the models to be loaded onto the GPUs (similar to a warm-up period).