

Unplatform AI
Open-source AI video and image studio. Cloud or local, no fees.
About
Unplatform AI is a free, open-source desktop app for making AI videos and images on Mac, Windows and Linux. Use cloud models with your own OpenRouter key and see the exact price before every generation, or run local models on your own GPU through ComfyUI. The app adds nothing on top, needs no account, keeps your keys in the system keychain and your files on your disk.
Generate video with Wan and Seedance, with start and end frames, reference images, video and audio. Create images with FLUX, Seedream, Gemini and OpenAI models in a fast masonry gallery, then turn any image into a video reference in one click. Favorites, prompt reuse, batch export and live prices are built in.
Local models: install the Unplatform AI node pack in ComfyUI, save a workflow, and it shows up as a model in the app. Cloud and local results share one history. The pack includes 9 ready workflows (19 models): FLUX.2 Klein, Qwen 2.1 (with head swap and body swap), Krea 2, Ideogram 4, Z-Image, MOSS Image, Wan 2.2, LTX 2.5 and MiniMax H3 Video / Image. Each comes with a guide listing its VRAM and RAM needs and a prompt cheat sheet, and ComfyUI offers to download the missing models when you open it.
Coming next: a built-in upscaler for photos and videos, trimming and cropping, and splitting a video’s audio into separate layers.
ComfyUI node pack: github.com/jefoce/unplatform-ai-provider-node



















Smooth launch video dude, and nice UI!