<h3 id="splashed-mix-v5-final-release-bhewb6s5f">Splashed Mix V5 – Final Release</h3>This is it — the final version of Splashed Mix.After five iterations of merging, testing, and refining, V5 pushes SDXL as far as it can go. It’s stronger in every way: richer textures, deeper prompt understanding, and a more intuitive feel for both art and photography. This version thinks differently — it’s not just an upgrade, it’s the sum of everything that came before.The settings remain the same: 8–10 steps, LCM Exponential, CFG 1.0 — to match the results shown, stick to these.V5 is a swansong — the last checkpoint in a long journey. Hope it inspires you to create something unforgettable.

Splashed Mix DMD

V4Important stuff to know:<ul><li>CFG: 1–1.5</li><li>Sampler: LCM using Exponential, AYS, or Beta</li><li>Steps: 8–10</li><li>Aspect Ratios: Most should work without distortion</li><li>Stick to these settings if you want results like the sample images—V4 doesn’t behave like SDXL or other base models, and it’s not meant to.</li></ul>

splashedMixDMD_v40.safetensors

v4.0

v5.0

v3.0

v2.0

v1.0

Unleash Unprecedented Creativity with Our Model v2.0Welcome to the next level of image creation with v2.0, our latest and greatest model yet. Building on the feedback from our amazing users, this version brings even more to the table. It's fast, versatile, and capable of generating stunningly photorealistic images from any prompt, whether it's a single word or a detailed narrative. The DMD integration ensures lightning-fast results in just 8 steps.Settings:LCM Exponential8 steps1 CFGWhat's New in v2.0?Vibrant Colors: Experience deeper and more dynamic hues.Enhanced Detail: Achieve an unprecedented level of realism.Improved Prompt Accuracy: Enjoy better fidelity to your creative visions.Greater Creativity: Now even closer to MidJourney in terms of artistic output.Not only does v2.0 maintain the photorealism you've come to love, but it also matches MidJourney's creative prowess. With the power of DMD, it transforms XL into a true artistic powerhouse.All sample images shared (100+ and counting) are direct outputs—no upscaling or edits—demonstrating the true potential of our model. Try using a MidJourney prompt and see which one inspires you the most.

About this version

V4

Important stuff to know:

CFG: 1–1.5

Sampler: LCM using Exponential, AYS, or Beta

Steps: 8–10

Aspect Ratios: Most should work without distortion

Stick to these settings if you want results like the sample images—V4 doesn’t behave like SDXL or other base models, and it’s not meant to.

Files with SHA256: 5d2cd441aac236601e6325c9b17a828265e53ba5b2050c52096e3ce35ada6538