Valentina is an original fictional character by dearai studio. This LoRA helps you create pictures of the same character in different outfits, expressions and settings.
Base model: Tongyi-MAI/Z-Image-Turbo
Trigger word: zxvalentina
Tested LoRA strength: 1.0
Include zxvalentina in your prompt, then describe the picture you want.
Example:
zxvalentina woman, smiling, wearing a green dress, standing on a balcony, warm evening light, realistic photograph
The example images were generated with this character LoRA, without additional LoRAs or a finishing pass.
Tested settings: 8 steps, flowmatch sampler, guidance 1 in AI Toolkit, resolution 1024 × 1344. Settings may differ in other applications.
Known limitations: Small facial marks and jewellery can vary. Hands, feet and written text may need checking. Scene details may not follow every instruction exactly.
Unload the LoRA when switching to another character; removing the trigger word alone may still retain Valentina’s appearance.
Explore our character studio: https://dearai.live
Description
First Z-Image Turbo release of Valentina.
Trained on 45 image-and-caption pairs for 2,500 steps. Selected after comparing five saved checkpoints.
Use with Tongyi-MAI/Z-Image-Turbo.
Trigger word: zxvalentina
Tested LoRA strength: 1.0
Includes eight selected example images showing different expressions, outfits and settings.
Small facial marks and jewellery can vary. Check hands, feet and generated text before using an image.