Krea is a multi-modal generative AI platform exposing 64+ state-of-the-art models (Flux, Veo 3.1, Kling, Hailuo, Wan, Runway, Ideogram, Gemini, and proprietary Krea 2) through a unified web app and API. It supports text-to-image, realtime canvas rendering (<50ms), text-to-video, image-to-video, motion transfer, text-to-3D, image-to-3D, LoRA fine-tuning (image and video), generative editing, frame interpolation up to 120fps, and upscaling to 22K px via Topaz and in-house models. The platform includes a node-based workflow automation engine, an asset manager, and a commercial API available for enterprise and developer integrations. Native 4K image generation is supported, with inference speeds of ~3s for a 1024px Flux FP16 image.