RHTV

RHTV is an infinite canvas tool produced by RunningHub. Its selling point is that it embeds the agent directly into the canvas instead of opening a dialog box next to it.

The difference in actual use is this: you use natural language to give instructions on the canvas, and it completes the steps of requirement analysis, scripting, generation, and synthesis. The intermediate products are left on the canvas and can be changed at any time, without having to describe the context again. Officially, this is called real-person machine collaboration, and is used to compare products that "look like canvases, but actually require re-feeding the context at every step."

The model is connected to Seedance 2.5, MiniMax, Wan 3.0, etc. The homepage also has quick entry points such as one-click movie making and hit lab, and there are many short drama templates.

RunningHub itself started out as a cloud-based ComfyUI, with a strong engineering background, and running batch tasks is more stable than pure front-end assembled products.

Video tools deal with images, but stories have to be found separately.TVLabWe have sorted out popular skits and common routines (counterattack, hidden identity, contract marriage...), and each inspiration card comes with vertical screen shots and screen prompts.