PoloX AI

PoloX AI is a web creative workspace with agent chat and an infinite canvas. It brings many image and video models together with reusable skills for text editing in images, object removal, background removal, storyboards and talking avatars.

PoloX AI is a browser-based creative workspace that combines a chat with agents and an infinite canvas for producing images and videos. The home screen offers a Classic mode, where a model, prompt, aspect ratio, resolution (1K, 2K or 4K) and background setting (auto, opaque or transparent) are chosen by hand, and an Agent mode that handles the work through conversation. A free generation option is shown beside a credit auto-confirm toggle. The platform bundles many third-party models in one place. The listed image models include GPT Image 2.5 Flare and Sunburst, Nano Banana 2.1, Nano Banana Pro, Seedream 5.0 Pro and Qwen Image 2.1, which also has a free variant. Video models include Wan 3.0, Seedance 2.0 and 2.5, Flux 3 and MiniMax H3, covering text-to-video, image-to-video and reference-to-video modes. Each model has its own page, so comparing options does not require leaving the workspace. A large part of the product is its Skills library: reusable creative workflows invoked by a slash name. Utility skills include a Product Hunt gallery builder, App Store graphics, sketch to image, an image text editor that rewrites only the text a user changes, annotated edits with numbered markers, object removal, and a layer splitter that exports transparent PNG layers. Video skills cover storyboard-driven multi-shot films, talking avatars with lip-sync in English or Chinese, and motion copy from a reference clip. A Fun category adds portrait styles and hairstyle filters, and users can create their own skills. Standalone utilities handle background removal and image-to-image editing. PoloX suits marketers, indie makers, designers and social creators who want several generation models and repeatable workflows without switching between separate apps. It sits between single-model generators and node-based canvas tools: model choice is broad and skills reduce prompt writing, though depth on any single task depends on the underlying model chosen.