Description: 🚀 Run this workflow for FREE on Runninghub: <a target="_blank" rel="ugc" href="https://www.runninghub.ai/post/2072151051746242562/?inviteCode=rh-v1497">https://www.runninghub.ai/post/2072151051746242562/?inviteCode=rh-v1497</a>(New users get 1,000 free credits + 100 daily login points—grab it while the promo lasts!) 🎥 Watch the full YouTube breakdown: <div data-youtube-video><iframe width="640" height="480" allowfullscreen="true" autoplay="false" disablekbcontrols="false" enableiframeapi="false" endtime="0" ivloadpolicy="0" loop="false" modestbranding="false" origin playlist src="https://www.youtube.com/embed/u6-vC8Ct09w" start="0"></iframe></div>💬 Join Discord for the Prompt Templates: <a target="_blank" rel="ugc" href="https://discord.gg/NmBhk9Vk8b">https://discord.gg/NmBhk9Vk8b</a>🌟 Overview Welcome to the 12-billion-parameter beast: Krea 2. While the raw version of Krea 2 is built for developers and AI "alchemists," this ComfyUI workflow makes its Turbo variant accessible to everyone. To solve the issue of annoying online LLM node installations, I have completely integrated a local Qwen 3.5 model to handle your prompt expansion automatically!⚙️ Core Features:<ul><li>Local LLM Prompt Structuring: Krea 2 requires highly structured prompts to shine. This workflow uses a built-in Qwen 3.5 node and a custom template to take your simple ideas and automatically expand them into Krea-friendly English prompts locally.</li><li>3-Pass Sampling Architecture: Ensures high-quality generation with precise step control.</li><li>Klein Upscaler Integration: Directly hooks into a Klein upscaler at the end of the pipeline for crisp, high-definition final outputs.</li></ul>💡 Usage Tip: Just set your dimensions, type a simple prompt in the designated node, and let the Qwen LLM do the heavy lifting.

Files with SHA256: dfa9d823d88e661f18111d70e8eaae492e49076e1375a163026eb8996cd47707