A realistic layer mix of several schnell models (shuttle 3.1, shuttle 3.0, schnell base) picking the best layers from each to produce an aesthetic model.<a target="_blank" rel="ugc" href="http://v5.photo">v5.1 (photo) - More photographic fine tuning</a>v5.0 -adds in yet more fine tuningv4.0 - Not publishedv3.3 - adds in some minor fine-tuningDon't forget to tip! Models like these take a lot of work to wrangle. Tip me some buzz!<h2 id="nag!-3491d86zb">NAG! </h2>I highly recommend trying NAG with this checkpoint <a target="_blank" rel="ugc" href="https://github.com/ChenDarYen/ComfyUI-NAG?tab=readme-ov-file">https://github.com/ChenDarYen/ComfyUI-NAG?tab=readme-ov-file</a> It gives you back control over negative prompts and can really increase the realism.<h2 id="recommended-settings-1tdcf4fml">Recommended settings</h2>Euler - Simple scheduler - 4 steps - CFG 1.0, using more steps = more details.This checkpoint is capable of doing 2 Megapixel images without issue. Suggested Resolutions:1MPix1:1 - 1024 x 1024 4:3 - 1152 x 896 16:9 - 1344 x 7681.5MPix1:1 - 1280 x 1280 4:3 - 1440 x 1120 16:9 - 1680 x 960 21:9 - 1920 x 8002Mpix 1:1 - 1408 x 1408 4:3 - 1664 x 1216 16:9 - 1920 x 1088<h2 id="recommended-models-by-gpu-vram-gqc5xsy40">Recommended models by GPU VRAM</h2>24Gb - FP8 / Q8 GGUF... or anything smaller16Gb - FP8 / Q8 GGUF... or anything smaller12Gb - FP8 / Q6_K GGUF / Q5_1 GGUF8Gb - Q5_1 GGUF / Q4_1 GGUF / Q4_K GGUF6Gb - Q3 GGUF / Q3K GGUF

schnellMODE (FLUX.S) - FP8, FP16, GGUF, NF4

This one may suffer from diagonal stair stepping on hair, if you find that, increase the steps or change the seed.

Files with SHA256: d82e783087d8fa7cc13866756ad4af610b9ab72b86f8d5dbe640d516077aaacb