A new character for the Cast. Black hair, striking blue eyes, magnetic presence — built to hold across scenes and styles.
Specs
Trigger word:
Z_M4rt1n4_ZTBase model: Z-Image Turbo
Recommended weight: 1.0 — sweet spot
Working range: 0.8 → 1.2 (best zone)
Notes
Stress-tested across a wide range of prompts and contexts — character holds, flexibility is there.
Tested solo and combined with style LoRAs I run regularly on Z-Image Turbo — she keeps her identity, no fight with the style.
All preview images here were generated with Martina alone, no style LoRA stacked. What you see is the LoRA on its own.
Tip on complex scenes: on busier compositions Z-Image Turbo can occasionally push anatomy off (extra limbs, doubled features — a known base-model behavior on complex prompts, not specific to this LoRA). If it happens: bump steps to 10 and drop LoRA weight to 0.8 — usually cleans it up.
If you cook something nice with her, drop it in the gallery.
Martina is the fifth drop of the ShinobiSat Cast — a growing series of photoreal character LoRAs. More coming.
— ShinobiSat
Description
Specs
Trigger word:
Z_M4rt1n4_ZTBase model: Z-Image Turbo
Recommended weight: 1.0 — sweet spot
Working range: 0.8 → 1.2 (best zone)
Comments (2)
very nice model, beautifil character and stunning face and captivating gaze. Great job, thanks for sharing 👍😍
Thanks! I'm glad you like it.
It’s a model I had trouble training—I was almost about to give up on it—but then, on the last attempt, it turned out really well.