<h1 id="krea-2-turbo-8-step-distillation-lora-(svd-extract)">Krea-2 Turbo 8-Step Distillation LoRA (SVD Extract)</h1>Makes Krea-2 Raw behave like Krea-2 Turbo (8-step, CFG-free) without downloading the 24 GB Turbo checkpoint. This is an experimental extraction, not an official Krea release.How to use<ol><li>Load the Krea-2-Raw transformer (not Turbo) in ComfyUI or diffusers.</li><li>Apply one LoRA to the diffusion transformer.</li><li>Use 8 steps, CFG 0–1, and timestep shift mu 1.15.</li></ol>Files<ul><li>krea2_turbo_distill_r64.safetensors: rank 64, ~0.47 GB. Smallest, rougher fit (use if VRAM or disk is tight)</li><li>krea2_turbo_distill_r128.safetensors: rank 128, ~0.94 GB. Recommended start</li><li>krea2_turbo_distill_r256.safetensors: rank 256, ~1.87 GB. Closest fit</li></ul>Notes<ul><li>This is an approximation. It recovers part of the Raw→Turbo shift, so it won't match native Turbo pixel for pixel.</li><li>Evaluate at the Turbo settings above.</li><li>Keys use the ComfyUI Krea2 convention (<code>diffusion_model.blocks.0.attn.wq.lora_A.weight</code>). Alpha = rank.</li><li>Method: truncated SVD of (W_turbo − W_raw) on each 2D layer, from krea/Krea-2-Raw and krea/Krea-2-Turbo.</li></ul>License: derived from Krea-2 weights, so it inherits the Krea-2 community license. See Krea's licensing for commercial terms.Krea 2 technical report: https://www.krea.ai/blog/krea-2-technical-report

Krea-2 Turbo 8-Step Distillation LoRA (SVD Extract)

<ul><li>r64: ~0.47 GB, smallest, rougher fit</li><li>r128: ~0.94 GB, recommended starting point</li><li>r256: ~1.87 GB, closest fit, largest file</li></ul>

Files with SHA256: ca178303453e1152d0e08a36240509b4711c3ccb7596915a20e910c973dd42ae