Krea 2 Turbo Text-to-Image
Krea 2 Turbo is a newly open-sourced, high-performance text-to-image model for fast, high-quality local generation in ComfyUI. It supports FP8 and GGUF formats to cut VRAM use — a strong open Stable Diffusion alternative.
✓ 100% free — download the workflow and run it locally.Requirements
- GPU: NVIDIA RTX 30XX / 40XX / 50XX (FP8 compatible)
- VRAM: 8 GB minimum
- OS: Windows
- Storage: 20 GB+ free
Model downloads
- Krea 2 Turbo GGUF (→ models/unet/) https://huggingface.co/vantagewithai/Krea-2-Turbo-GGUF/tree/main
- Krea 2 Turbo FP8 (→ models/diffusion_models/) https://huggingface.co/AlperKTS/Krea2_FP8/tree/main
- qwen3vl_4b_fp8 text encoder https://huggingface.co/Comfy-Org/Qwen3-VL/tree/main/text_encoders
- qwen_image_vae https://huggingface.co/Comfy-Org/Qwen-Image_ComfyUI/tree/main/split_files/vae
Custom nodes
- ComfyUI Manager https://github.com/ltdrdata/ComfyUI-Manager
Setup steps
- Install ComfyUI Manager and its requirements.
- Place each model in its ComfyUI folder (unet, diffusion_models, text_encoders, vae).
- Load the workflow below and run it with a text prompt.
Notes & tips
- Use ComfyUI Manager's "Install missing custom nodes" if a node fails to load.
- Prefer one-click? A ready-made Krea 2 installer is available.
Want the 1-click version?
Skip the manual model downloads and node installs — get a ready-to-run installer, or unlock all 75+ with Local Lab Pro.