Krea-R Turbo is a quantized text-to-image model designed to run on consumer GPUs with at least 8 GB of VRAM. It is distributed in the GGUF format and incorporates two RealRebelAI style LoRAs directly into the transformer, enabling a blended visual style within a single file and eliminating the need to stack additional components at runtime.
This model is a derivative of krea/Krea-2-Turbo, having undergone LoRA merging before being quantized. It is built specifically for use with the ComfyUI-GGUF loader, but requires a krea2-aware fork—RealRebelAI/ComfyUI-GGUF_KREA-2—due to its custom architecture. Attempting to load it with standard GGUF nodes or loaders that do not recognize the 'krea2' architecture will result in failure. The model is intended for use with English text prompts and is suitable for those seeking a text-to-image generation solution that leverages quantized weights and integrated style LoRAs.
Krea-R Turbo is released under the krea-2-community-license. Its distribution on Hugging Face provides access to the model card and related files, supporting experimentation and deployment by users familiar with the required hardware and software environment.
Krea R Turbo is an Image generation product. It allows developers and creators to generate images from text prompts using a local, open-source model. Krea R Turbo is an open-source project aimed at AI researchers and creative developers. The project is open source (Apache-2.0). It runs on the web and the command line, and it can be self-hosted.
realrebelai builds and maintains Krea R Turbo, and the product first shipped in 2024. The project is developed in the open on GitHub with 10 commits in the last 90 days. Among its 4 catalogued features are text-to-image generation, quantized model, and Local GPU inference.
Latest indexed changes and source events
realrebelai/Krea-R-Turbo verified by the PulseGate indexer
Other apps tracked under the same category.
