PhotorealTouch — realistic skin and photographic detail across multiple base models PhotorealTouch is a family of LoRAs designed to add natural skin texture, pores, fine vellus hair and more believable light response without forcing a new identity or art style. It can be used for portraits, fashion, characters and full-body photography, with both adult women and adult men. Choose the version that matches your base model • Z-IMAGE V1.0 — for Z-Image / Turbo workflows • SDXL v3 — for SDXL 1.0 checkpoints; start at 0.60, use 0.45–0.65 for subtle texture and 0.70–0.85 for stronger detail • SDXL v2.0 and SDXL V1 — earlier SDXL releases • PONY v1.0 — for Pony-based checkpoints • Illustrious — for Illustrious-based checkpoints • SD 1.5 version — for Stable Diffusion 1.5 checkpoints • FLUX — for compatible FLUX workflows Always select the matching version before downloading or generating. Recommended weights can differ between architectures. What it helps with • Natural pores and skin micro-texture • Less plastic-looking or airbrushed skin • More convincing facial detail and soft lighting • Photographic realism without imposing a fixed identity SDXL v3: tested workflow No trigger word is required. Useful prompt helpers include: realistic skin, natural pores, fine skin texture, raw photo. On already detailed or stylized checkpoints, start with a lower weight. See the reproducible SDXL comparison at weights 0.4, 0.6 and 0.8, including a no-LoRA control: https://civarchive.com/articles/35099/what-changes-when-lora-weight-changes-can-you-actually-see-a-difference Tips Compare the same prompt and seed with and without the LoRA, then adjust the weight before changing the checkpoint. Avoid stacking several strong detail enhancers. If PhotorealTouch improves your result, please publish it in the gallery and leave a review. User examples help document how the LoRA behaves across different checkpoints, lighting conditions and subjects.