<h1 id="darkbeast-krea-2-turbo-merge">DarkBeast Krea 2 Turbo Merge</h1>All credit goes to the original authors of <a rel="ugc" href="https://www.krea.ai/">Krea</a> and the wizards at @AiMetatron and their already perfect <a rel="ugc" href="https://civarchive.com/models/2242173/dark-beast-or-int8-convrot-2-or-krea2-aggressive-edition?modelVersionId=3078453">Dark Beast | 黑兽 INT8 Convrot 2倍速 | Krea2 Aggressive Edition</a>I wanted to try a model merge so I took the OG Krea 2 Turbo and blended in the Dark Beast. I am happy with the results so I will share here, but I encourage you to download the original Dark Beast and Krea 2 Turbo models for your creations as my merge may not be to everyone's liking.The results?A custom Krea 2 Turbo merge combining the speed, prompt adherence, and low-step performance of Krea 2 Turbo v1.0 with the darker, more dramatic visual character of DarkBeast Krea 2.The goal of this merge was not to create a simple 50/50 blend. Instead, the model preserves most of Krea 2 Turbo’s core conditioning and early transformer structure, while gradually introducing more DarkBeast influence through the middle and later transformer blocks.The result retains excellent performance at 8 steps, while adding richer contrast, moodier lighting, more detailed textures, and a stronger cinematic or dark-fantasy aesthetic.<h2 id="merge-structure">Merge Structure</h2>Model 1: Krea 2 Turbo v1.0 Model 2: DarkBeast Krea 2The merge uses Krea 2 Turbo as the primary structural base.Key timestep and projection layers remain approximately 90% Krea 2 Turbo to preserve distilled low-step behaviour. Text-conditioning layers remain mostly Turbo-biased for reliable prompt adherence.DarkBeast influence increases gradually across the transformer stack:<ul><li>Blocks 0–6: 25% DarkBeast</li><li>Blocks 7–13: 35% DarkBeast</li><li>Blocks 14–20: 40% DarkBeast</li><li>Blocks 21–27: 45% DarkBeast</li></ul>The final output section returns to a stronger Krea 2 Turbo weighting for stability.<h2 id="merge-ratios">Merge Ratios</h2><pre><code>first. 0.80
tmlp. 0.90
txtmlp. 0.70
tproj. 0.90

txtfusion.layerwise_blocks.0. 0.70
txtfusion.layerwise_blocks.1. 0.70
txtfusion.projector. 0.80
txtfusion.refiner_blocks.0. 0.65
txtfusion.refiner_blocks.1. 0.65

blocks.0 through blocks.6 0.75
blocks.7 through blocks.13 0.65
blocks.14 through blocks.20 0.60
blocks.21 through blocks.27 0.55

last. 0.70</code></pre>Ratios represent the amount of Model 1, Krea 2 Turbo v1.0. The remaining percentage comes from DarkBeast Krea 2.<h2 id="recommended-settings">Recommended Settings</h2>The model performs well using standard Krea 2 Turbo workflows.Suggested starting point:<pre><code>Steps: 8
CFG: 1.0
VAE: qwen_image_vae.safetensors
Text encoder: qwen3vl_4b_fp8_scaled.safetensors
Sampler: er_sde
Scheduler: sigmoid_offset
You can also use euler/simple</code></pre>The model can also be tested at 9 or 10 steps for additional refinement, although 8 steps already produces strong results. Too high add unwanted noise.<h2 id="strengths">Strengths</h2><ul><li>Strong performance at low step counts</li><li>Good prompt adherence</li><li>Cinematic and dramatic lighting</li><li>Rich contrast and atmospheric rendering</li><li>Detailed faces, clothing, textures, and environments</li><li>Excellent skin texture</li><li>Works well for NSFW or SFW</li><li>Particularly suitable for dark fantasy, gothic, cinematic, fashion, portrait, and dramatic photographic imagery</li></ul><h2 id="notes">Notes</h2>This is an experimental model merge and may behave differently from either parent model depending on the prompt and workflow.It was created with ComfyUI’s native <code>ModelMergeKrea2</code> node and saved as a standalone Krea 2 diffusion model.Please do not share or re-merge this model without giving credit to @AiMetatron

BF95-DarkBeast_Krea_Turbo_8-Step_Cinematic_Krea2_merge

V1.0 - test merge. See description for more info.

Files with SHA256: d64d90d0e0a55490f98b680ffa82e96e5ae4596539ab8dda1b2927f8a7c2be20