EVA — AI Influencer LoRA (Z-Image Based)

EVA is a stylized AI influencer designed to create eye-catching, social media–ready content with a consistent and recognizable look.

She leans into a soft “baddie” aesthetic — expressive eyes, clean skin, and a confident vibe that fits perfectly into modern Instagram and TikTok-style visuals.

Trained on the Z-Image Base model, EVA performs especially well with Z-Image Turbo, delivering sharp details and natural-looking results while keeping her identity stable across different scenes.

🔑 Trigger Words:
Baddie1, Double Bun, Blond Hair

⚙️ Training Details:

• Base model: Z-Image

• Steps: 3000

✨ What makes EVA stand out:

• Strong facial consistency across generations

• Naturally photogenic look (works great for “camera roll” style shots)

• Adapts well to different lighting and environments

• Perfect balance between realism and stylization

📸 Best use cases:

• AI influencer content

• Instagram / TikTok style posts

• Selfies, mirror shots, lifestyle photos

• Casual or aesthetic daily scenes

📊 Training Data:
EVA was trained on a carefully curated dataset of 40 high-quality images, featuring diverse lighting conditions and face angles. This allows the model to stay consistent while adapting naturally to different scenes and environments.

🧪 Example Prompts:

1. baddie1, selfie, blond hair, double bun, holding phone, slight smirk, messy bedroom, unmade bed, clothes on chair, soft window daylight, casual candid vibe

2. baddie1, selfie, blond hair, double bun, in bed under blanket, only face visible, phone light illuminating face, dark room, intimate atmosphere

3. baddie1 chest-up portrait, blond hair, double bun, laughing naturally, friends blurred in background, outdoor cafe, golden hour sunlight, candid lifestyle shot

📌 Notes:
EVA works best when you lean into her identity — using the trigger words helps maintain her signature look, but she’s flexible enough to experiment with different outfits, moods, and settings.