Multi-model image and video generation
Use Krea and third-party models for text-to-image, image-to-image, text-to-video, and image-to-video work in one interface.
Tool profile · Images
Krea combines its own and third-party visual models in one workspace for image, video, 3D, real-time generation, generative editing, upscaling, and LoRA fine-tuning.
01 / Overview
02 / Core features
Use Krea and third-party models for text-to-image, image-to-image, text-to-video, and image-to-video work in one interface.
Continuously reshape results with brushes, primitives, prompts, camera input, or screen streams.
Upscale and reconstruct images, enhance video, and interpolate frames with plan-dependent models and limits.
Generate 3D objects and fine-tune reusable LoRAs for people, products, or visual styles.
Keep sessions across tools and continue assets through generation, editing, and enhancement workflows.
03 / Use cases
Explore composition, lighting, style, and mood in real time before selecting a direction to refine.
Create product imagery, posters, and social assets, then refine them with editing and enhancement tools.
Compare video models and test image-to-video, motion direction, and short-shot atmosphere.
Improve detail, resolution, and frame rate for old media, low-resolution assets, renders, or video.
04 / Platforms and languages
05 / Pricing and quotas
Currently shows 100 compute units per day with real-time models and limited image, video, 3D, lipsync, enhancement, and LoRA access.
Currently shows 5,000 compute units monthly plus commercial use, full image/3D/lipsync access, LoRA fine-tuning, and selected video models.
Currently shows 20,000 compute units monthly with all video models, Nodes and Apps workflows, and higher concurrency.
Currently shows 60,000 compute units monthly with unlimited LoRA fine-tuning, expanded concurrency, and relaxed generation on selected models.