Original Model: Tifa Lockhart [Krea2] - Improved Likeness on CivitAI

Original Creator: LatentHeart on CivitAI

Recommended Settings:

• LoRA Strength: 1.0

• Trigger Word: TifaLockhart

The sample images were generated with this workflow .

Original Creator Description:

"💡 Make sure you are using the trigger word in your prompt: TifaLockhart
This is because the model already has Tifa baked in and you might not notice that the LoRA is not triggered correctly.

This LoRA is designed to improve Tifa Lockhart's likeness, not to introduce the character into the model.

Krea2 already has a strong understanding of Tifa, but this LoRA refines her facial features, proportions, and overall identity, making her more consistent and recognizable across a wide variety of generations while preserving the flexibility of the base model.

Because it focuses on improving character fidelity rather than imposing a specific visual style, it works well with virtually any aesthetic supported by Krea2, including:

• Realistic photography

• 3D renders

• Anime and manga

• Digital painting

• Concept art

• Stylized illustrations

• Fantasy artwork

• And other custom styles

Simply combine the LoRA with your preferred style prompts to achieve the look you want while maintaining a stronger resemblance to Tifa."

Support Original Creator:

Ko-fi: LatentHeart on Ko-fi

DeviantArt: LatentHeart on DeviantArt