Imaginode

Imaginode is an image and video generator built as a canvas rather than a chat box. You drop nodes for text, image, video, voice, camera and references, wire them together, and rerun one step without redoing the rest. Every node keeps its own history, so nothing you generated disappears. Forty eight models sit behind a single credit balance: Flux, Imagen 4, Seedream, Recraft and GPT Image for stills, Kling, Seedance, Veo 3.1 with audio, MiniMax, Wan and PixVerse for video, plus ElevenLabs for voice over. You never open an account with each provider, and you never juggle several subscriptions. A concrete run looks like this. Write a prompt, generate four variations, keep one, send it to a video node as the first frame, add a camera move, then attach a voice track in the same board. Change the prompt on the image node and only the steps downstream rerun. Each node shows what it will cost before it runs, and a failed generation refunds its credits automatically. There is also a public REST API and an MCP server for driving the same models from your own code. The interface is available in French, English, Spanish, German and Portuguese, and boards can be shared for real time collaboration.