What problem does it solve? Running AI generations one at a time makes multi-shot scenes, ad campaigns, and storyboards hard to coordinate. This Skill explains how to use Higgsfield Canvas, a node-based infinite board, to wire prompts, references, and generators into one re-runnable pipeline. ## Core Features & Use Cases - Node-Based Pipelines: Connect Prompt, Image generator, Video generator, Voice/audio, LLM assistant, Style/motion, and Render output nodes with type-aware edges. - Seven Models on One Board: Run Soul 2.0, Seedance 2.0, Kling 3.0, Wan 2.7, Veo 3.1, GPT Image 2.0, and Nano Banana Pro together in a single graph. - Named Patterns & Templates: Start from Simple Seedance, Extend Video, Image Edit, StoryBoard With Elements, or Long Video fan-out layouts, then save any board as a reusable template. - Use Case: Fan one product reference into several shot prompts, generate each shot independently, and assemble a multi-shot ad campaign while paying only for the generations you execute. ## Quick Start Ask the assistant to design a Higgsfield Canvas pipeline that turns one product reference image into a multi-shot video ad using the Long Video fan-out pattern.
