<h2 id="8-steps-cfg-1-a-picture-in-about-six-seconds-on-a-5090."><strong>8 steps, CFG 1, a picture in about six seconds on a 5090.</strong></h2><p><span style="color:rgb(250, 176, 5)">(Text to Image + Image to Image)</span></p><p>Two ways to run it, chosen with one number. There is a single blue INPUT box near the left:</p><p>1 = text to image. Type a prompt, press Run.</p><p>2 = image to image. Load your own picture, press Run.</p><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/rst_h25FAXM" start="0"></iframe></div><p>Two samplers sit behind that box with the denoise already correct for whichever you pick, so there</p><p>is one knob and not two. The half you did not choose never runs — nothing to mute, nothing wasted.</p><p></p><p>At CFG 1 the negative prompt does nothing at all. Not a little — nothing. That is what CFG 1 means, and it is why the box is left empty in this file. If you want a negative prompt to work, take the Extended file instead.</p><h3 id="what-it-runs-on"><strong>WHAT IT RUNS ON</strong></h3><p>Our own Krea 2 build, and it is free — every build on that page opens for any signed-in member, no code and no card. You also need the Krea 2 text encoder and VAE from the official Comfy-Org</p><p>repository, and the ComfyUI Impact Pack from ComfyUI Manager (that is the switch, and it is the only extra thing this file needs).</p><h3 id="watch-the-vae:"><strong>Watch the VAE:</strong></h3><p>it is qwen_image_vae, NOT wan_2.1_vae. Use the wrong one and nothing warns you — no</p><p>error, no red border. It runs every step, decodes, saves a picture, and the picture is coloured</p><p>static.</p><h3 id="there-are-three-of-these-workflows"><strong>THERE ARE THREE OF THESE WORKFLOWS</strong></h3><p><strong>Turbo</strong></p><p>(8 steps, CFG 1, ~6s, no negative prompt),</p><p><strong>Extended</strong></p><p>(16 steps, CFG 2, ~22s, negative works),</p><p><strong>Base</strong></p><p>(20 steps, CFG 3, ~46s, negative works).</p><p>They are genuinely different files, not three presets of one. Start with Turbo if you just want to see something.</p><h3 id="everything"><strong>Everything</strong></h3><p>Links to all three files, the model, the folder each file goes in, and the settings: <a target="_blank" rel="ugc" href="https://forgebun.com/go/civit_krea2_workflows">grab it here</a></p><h3 id="the-model-on-its-own:"><strong>The model on its own:</strong></h3><p><a target="_blank" rel="ugc" href="https://forgebun.com/go/pro_models">Is here</a></p><p>Workflow, settings and write-up are mine. Krea 2, its text encoder and VAE are Comfy-Org's. The</p><p>switch node is from the ComfyUI Impact Pack by ltdrdata.</p>

Krea 2 Turbo T2I + I2I ComfyUI workflow

<p><strong>v2 fixes the INPUT switch.</strong></p><p>It would not move from 1 to 2, so the picture-to-image side of the workflow was out of reach. That was a fault in the file, not in your ComfyUI. Nothing else has changed: same nodes, same settings, same models.</p><p>Thanks to everyone who reported it.</p>

Files with SHA256: 89d685e28cfaf7eb8e13e13ddf95e52a577bd4793d919d2906b4dd344f5e1504