Wind Comic
Wind Comic is a multi-agent AI pipeline that turns one line of text into a finished short-form drama, script, character-consistent storyboards, voice acting, music and a rendered video, using specialist agent roles for writing, directing, production and editing.
Wind Comic runs a pipeline of specialist agent roles, writer, director, producer, character designer, storyboard artist, cameo locker, lip-sync and editor, that hand off to each other under strict consistency contracts, rather than relying on one large model to do everything at once. Starting from a single sentence, or from a longer novel that it can split into episodes, it produces a script, a character bible, multi-shot cinematic storyboards, voice-acted lines, background music and mouth-synced characters, assembled into a finished video file.
It is provider-agnostic for both language and media models, drawing on OpenAI or Claude for text, and MiniMax, Midjourney, Veo, Sora, fal or ComfyUI for images and video, with a language-model gateway that falls back from one provider to another with no code change. A real-time collaborative timeline lets a team edit the result together.
It is written in TypeScript with Next.js and released under the MIT license.