Meet Aerith — the flower girl from the slums, the one everybody remembers. This LoRA is a fan tribute built around her likeness: her face, her warm and slightly mischievous presence, that gentle look that makes a scene feel calm.One thing up front, because it changes how you prompt her: this is a character-identity LoRA. It reproduces who she is, not a fixed costume. Her iconic pink dress, the red jacket and the hair ribbon were not part of the training set — the dataset is deliberately varied everyday clothing and locations, so her identity holds up in any setting. If you want the signature look, describe it in your prompt; the LoRA keeps her face while you decide the wardrobe. Results are inspired by the character, not screen-accurate to the game.It is tuned for Z-Image Turbo, so you get a photorealistic Aerith at turbo speed: few steps, low CFG, fast iterations, strong identity. There are also versions for Illustrious and FLUX.2 Klein 9B. Have fun with her — and please share what you make.<h2></h2><hr /><h2 id="recommended-settings">Recommended Settings</h2><ul><li>Trigger word: aerithtailor (put it at the very start of your prompt)</li><li>Strength: 1.0 (useful range 0.85 – 1.0)</li></ul><h3 id="z-image-turbo">Z-Image-Turbo</h3><ul><li>Steps: 9</li><li>CFG: 1.0</li><li>Sampler: res_multistep</li><li>Scheduler: simple</li><li>Shift: 3.0 (via a ModelSamplingAuraFlow node — Z-Image / AuraFlow specific)</li><li>Negative prompt: (empty)</li><li>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.</li></ul><h3 id="illustrious">Illustrious</h3><ul><li>Steps: 30</li><li>CFG: 4.0-5.0</li><li>Sampler: dpmpp_2m, dpmpp_3m_sde</li><li>Scheduler: karras</li><li>Models: Realism Illustrious By Stable Yogi, Ultra Realistic By Stable Yogi, ramthrustsNSFWPINK_alchemyMix176, perfectdeliberate v70, iLustMix</li></ul><h3 id="flux.2-klein-9b">Flux.2 Klein 9B</h3><ul><li>Steps: 25-30</li><li>CFG: 1.0</li><li>Guidance: 3.0-4.0</li><li>Sampler: euler / res_multistep</li><li>Scheduler: simple</li><li>Negative prompt: (empty)</li><li>Base model: Flux.2 Klein 9B + FP8, Jib Mix Klein,</li></ul><hr /><h2 id="prompt-examples">Prompt Examples</h2><h3 id="portrait-prompt:">Portrait prompt:</h3><pre><code>A photo of aerithtailor, a woman. A half-body portrait, three-quarter view, warm tones, black background, shallow depth of field, photorealistic, professional photography.</code></pre>Meaning: just her, cleanly lit, nothing competing with the face. Use this to verify the likenessbefore you build a bigger scene.<hr /><h2 id="training-details">Training Details</h2><ul><li>Source: curated character dataset (portraits, full-body, scenes, outside, inside, ...)</li><li>Dataset size: 89 images</li><li>Resolution: 768 px</li></ul><hr /><h2 id="notes">Notes</h2><ul><li>Fan service / disclaimer: This is a fan-made tribute LoRA. It is not affiliated with, endorsed by, or connected to Square Enix or any Final Fantasy rights holder. The character and all trademarks belong to their respective owners. For personal, non-commercial fan use.</li><li>Identity, not a costume: the pink dress, the red jacket and the hair ribbon were not in the training data. Describe the clothing you want in your prompt — the LoRA supplies the face.</li><li>Describe the clothing you want in your prompt — results are inspired by, not screen-accurate. Any accessory must be prompted and only makes sense in front / side views.</li></ul>

Files with SHA256: 3f9723ff392c2540817783b7845a399af3dd7ca3c0bbdce4521b9c3e49e6e89a