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

Additional merging and fine tuning. Still Apache license.

schnellmodeFLUXSFP8FP16_v5FP8.safetensors

v5 FP8

v5.1 (photo) Q4_K_S

v5.1 (photo) FP8

v5 Q4_K

V3.3 - Q5_K GGUF

V3.3 - Q4_K_S GGUF

V3.3 - Q8 GGUF

V3.3 - NF4

V3.3 - FP8

V3.1 (beta test) - FP8

V3.1 - NF4 (Beta)

V2 - FP16

FP16

v2 - Q5_1 GGUF

v2 - Q4_1 GGUF

V2 - Q3_K GGUF

V2 - Q8 GGUF

V2 - Q6_K GGUF

V2 - NF4

V2 - FP8

Q5.1 GGUF

Q3 GGUF

Q4 GGUF

Q6_K GGUF

Files with SHA256: 75f4d608f851c7f35b2e948680e0ba6de3836c11305aa6745d6c16e18ea53210