灵感皆可成型，创意无所不能

How Studio works

ChatPod Studio is an AI workbench for images, video, and website concepts on one infinite canvas. Type a brief in the composer, attach a reference, or pick a skill, then send. The same project keeps prompts, stills, motion boards, and annotations together so you iterate without splitting files across chat threads. Conversation in ChatPod stays for research and Q&A; this page is where visuals are actually made.

Typical work includes brand stills, social packs, ecommerce hero shots, product explainers, storyboard-driven clips, and landing-page directions. Failed or queued generations stay labeled. Demo or unavailable skills are not presented as live production. Opening a card or dropping a file places a hosted asset on the canvas instead of a disposable data URL, so references survive reload. Login is requested only when you enter the canvas or start a paid generation.

从自然语言或参考图进入画布，继续做图、视频或页面方案；项目可保存并对照版本。发布前请阅读 product overview、 terms 与 privacy notes。Operators choose which models, galleries, and storage backends are enabled; capability depends on that deployment, not on marketing copy. Do not treat generated drafts as final art direction or legal clearance. Empty recent-project rows stay empty until you save a canvas; inspiration cards only launch when you click them. Support and data requests go through the contact page, not through generation APIs.

Studio is built for repeatable visual work, not one-off chat screenshots. Keep references on the board, compare versions in the same project, and only publish after a human reviews likeness, brand, and claims. Website mode drafts structure and look; it is not a hosted CMS. Video mode explores pacing and camera language; it is not a finishing suite. If a model is disabled in your deployment, the control stays visible but cannot start a run. That honesty is intentional.