No Claude Motion on your plan? A 30-second Sume video from $1.875

Claude Motion is a beta on Team and Enterprise plans, per Anthropic's 2026-10-08 announcement, so a Free or individual account cannot use it yet. If you only need a 30-second video, Sume can generate one through its API for between $1.875 and $17.334 depending on the model and resolution. These are not the same product: Motion animates text and charts as code, while Sume's video models generate footage.

The announcement also says Docs, Slides and Design are out of beta on every plan, including Free. If your goal is a slide-style explainer, that may already cover you.

What a 30-second clip costs

Sume bills video per output second, and the rate is linear inside each model's valid duration range. The table multiplies each rate by 30 seconds.

Limits that decide the choice

Costs scale with the number of attempts. If you expect to regenerate once, double the figure before you commit a budget.

• wan-3.0 accepts 2 to 30 seconds per clip, so 30 seconds is one job.
• minimax-h3 accepts 5 to 15 seconds, so 30 seconds needs two clips plus a join. A Timeline render of the two clips is $0.10, which makes the total $2.35.
• seedance-2.5 accepts 4 to 30 seconds and is the highest-priced row here.
• These figures are Sume list prices computed as provider list x 1.25, default options, no extras.

Check live rates before you budget

Prices change, so read them at request time. Sume's video API lists models at GET /v1/videos/models, and the public catalog at GET /v1/catalog needs no key. Submit with POST /v1/videos and poll GET /v1/videos/{jobId} for the result.

For a first test, generate the cheapest allowed length instead of the full 30 seconds. At wan-3.0 480p a 2-second clip is 2 x $0.0625 = $0.125, which tells you whether the look is right before you spend $1.875 on the full-length version.

Which tool for which job

Pick Motion when every word and number must be exact and editable, because the output is code you can change. Anthropic gives a quarterly report as a 30-second explainer and a chart animation for a board deck as examples. Pick a video model when you need footage, a scene or a product shown in use, because a code animation of text and shapes cannot produce that. The Sume video docs cover the request fields and each model's valid duration range.

A generated clip may not render exact chart values. If numbers matter, put them in as captions or overlays afterwards. Sume's caption endpoint accepts authored cues at $0.20 per job for videos of 60 seconds or less, so the exact figures can be burned on top of the generated footage.