Lara Croft - Tomb Raider - Classic

Meet the classic Lara Croft — the fearless adventurer who defined a whole genre. This LoRA is a loving fan tribute that captures her likeness: her face, her features, that confident adventurer presence — tuned for Z-Image Turbo, so you get a photorealistic Lara at turbo speed (few steps, low CFG, fast iterations).

One thing up front, so you get the best results: this is a character-identity LoRA. It reproduces *who she is*, not a fixed costume — her classic in-game outfit was not part of the training set.

You decide what she wears by describing it in your prompt, from rugged expedition gear to an elegant portrait. Her identity stays locked in; the styling is yours.

Purely as a fan tribute. Have fun — and please share what you make.

---

Recommended Settings

• Trigger word: laracroftclassictailor (put it at the very start of your prompt)

• Strength: 1.0

• Steps: 9

• CFG: 1.0

• Sampler: res_multistep

• Scheduler: simple

• Shift: 3.0 (via a ModelSamplingAuraFlow node — Z-Image / AuraFlow specific)

• Negative prompt: (empty)

• Base model: Z-Image Turbo — a realistic Z-Image Turbo checkpoint like AnalogMadness (Z-Image Turbo) v2.0 or cyberrealistic Z-Image v0.5 gives the most photographic look.

---

Prompt Examples

Portrait prompt:

laracroftclassictailor, photorealistic close-up portrait, soft natural window light, 85mm

Meaning: the classic Lara in a clean, photoreal portrait — the fastest identity check.

Dress her yourself — the outfit comes from your prompt, not the LoRA:

laracroftclassictailor, full body, rugged expedition outfit, tank top and shorts, ancient stone temple, torchlight, dynamic pose

Meaning: the LoRA keeps her identity while you describe the clothing and scene (adventurer *style*, not screen-accurate to the game).

Note: if her identity or outfit doesn't always come through, reinforce the trigger with parentheses and a weight:

(laracroftclassictailor:1.1), full body, classic adventurer outfit, jungle waterfall

This boosts the LoRA's influence in the prompt.

---

Training Details

• Source: curated character dataset (portraits, full-body, action, scenes)

• Dataset size: 74 images

• Steps: 3000

• Rank / Alpha: 16 / 16

• Learning rate: 1e-4

• Trainer: ai-toolkit

• Base model: Z-Image Turbo

---

Notes

• Fan service / disclaimer: This is a fan-made tribute LoRA. It is not affiliated with, endorsed by, or connected to Crystal Dynamics, Eidos, Embracer Group, or any Tomb Raider / Lara Croft rights holder. The character and all trademarks belong to their respective owners. For personal, non-commercial fan use.

• Identity, not a costume: the classic in-game outfit and accessories were not in the training data.

• Describe the clothing you want in your prompt — results are inspired by, not screen-accurate. Any accessory (e.g. a medallion) must be prompted and only makes sense in front / side views.

• NSFW-capable.