Flux Text to Image with LoRA

SeaArt’s official model loader already supports FLUX tags and filtering, so we won’t repeat how to find FLUX Checkpoints and LoRA. Currently, FLUX supports multiple LoRA, but to keep the interface simple only one LoRA slot is available. FaeTastic Details and mjV6 are preloaded by default, each with a weight of 1. The compute cost for a single image is about 30 PT; when generating 2–4 images at once, the per‑image cost decreases, down to a minimum of 19 PT per image. For the best cost performance, we recommend increasing the output batch size.