Avatar script too long? The 4 to 60 second window and how to split it

The window

The Avatar videos docs say Sume accepts scripts and multi-scene plans when it estimates the video duration at 4 to 60 seconds inclusive. For longer scripts, make them shorter or split them into multiple jobs. In a video_inputs plan, a text voice takes script or input_text, not both; a silence voice needs a duration.

Two ways to split

• Scenes in one job: use video_inputs with an id per scene. The job keeps one avatar, and backgrounds resolve to one shared scene. Total must still be 60 s or less.
• Several jobs: send chapters as separate requests with separate Idempotency-Keys, then join them in your editor. The 4 to 60 s rule applies to each job.

Part one of a split script

curl -X POST https://api.sume.com/v1/avatar-1.0/talking-video \
  -H "Authorization: Bearer $SUME_API_KEY" \
  -H "Content-Type: application/json" \
  -H "Idempotency-Key: long-script-part-1" \
  -d '{"avatar_handle":"product_host","quality":"standard","script":"Part one of the walkthrough: what changed in this release and why it matters."}'

Estimate before you send

A spoken script runs roughly two words a second, so about 120 words is a minute; treat that as a planning rule, then check the cost with a dry run where the tool supports it.