A chart may look finished until someone spots the wrong number. The ability to correct that detail matters as much as the first animation.
Anthropic's Claude Motion guide describes short animations built as code, using text, charts, shapes, and images. Individual words, values, and timing remain editable. Its 8 October announcement presents that control as part of the product: make an animation, revise its details, and export an MP4.
As checked on 11 October, Motion is in beta for Team and Enterprise. The guide says Team has it enabled by default; an Enterprise owner must enable it. Animations use the plan's normal usage allowance, with longer or more complex work consuming more. Creating one is therefore subject to the account's access and limits.
Try the correction, too
My suggested test is a short explanation of a process you understand well. Use your own diagram or a handful of known values. Specify the audience, duration, and the one idea the viewer should leave with. Keep the first attempt small enough to inspect scene by scene.
Then change one detail. Replace a value, slow a transition, or correct a label. Watch the whole piece again. Does the explanation still make sense? Did the surrounding chart, narration, or emphasis need a corresponding change? An editable number does not automatically make the argument around it accurate.
That is the interesting design question here: how well does a generated explanation survive revision? The first result demonstrates generation. The second reveals whether the tool can become part of a working editing process.