YouTube Shorts series: 180-second episodes, make a season in batch
YouTube Shorts series group Shorts into seasons of episodes up to 180 seconds. Plan a season, keep one look, and batch the jobs on Sume.
A YouTube Shorts series is a show whose episodes are Shorts: portrait or square, at most 180 seconds each, grouped into seasons. You can build one with Sume by writing the season as a list of scripts, rendering each as an Avatar 1.0 talking video of 4 to 60 seconds, and keeping the same avatar and scene across every job, though Sume does not publish to YouTube or create the show.
The rules come from YouTube's Create shows from your videos page and its September 23, 2026 Shorts series blog post, both read on 2026-10-02. Sume's side comes from Generate avatar video and Timeline 1.0.
What does YouTube require for a Shorts series?
The blog says Shorts series is rolling out to creators across web, mobile and TV, and lets creators structure Shorts into seasons and episodes, with custom thumbnails and sequential playback. The help page adds the format limits. It says the show page needs consistent story or topic, a consistent look or feel and high-quality video and sound, and that videos with no visuals will probably not be considered a show.
Episodes are numbered by publish date, a show supports multiple seasons, and it can be serial, meant to be watched in order, or non-serial.
| Item | Rule on the YouTube page |
|---|---|
| Shape | Portrait or square only |
| Episode length | Maximum 180 seconds |
| Numbering | By publish date |
| Structure | Multiple seasons; serial or non-serial |
| Quality bar | Consistent topic and look, high-quality video and sound |
How does Sume fit those limits?
Avatar Video supports 1:1, 3:4, 9:16, 4:3 and 16:9, defaulting to 9:16, so 9:16 and 1:1 are the two that satisfy YouTube's shape rule. A script is accepted when Sume estimates the video at 4 to 60 seconds, so each episode stays well under 180 seconds; longer episodes need Timeline, which renders up to 1800 seconds.
Consistency is the part Sume helps with most. The current execution supports one resolved avatar per final video and expects scene backgrounds to resolve to one shared scene, so you reuse the same avatar_handle and scene prompt for every episode.
A season plan helps before any rendering. Give every episode a number, a one-sentence promise and a script length, and keep the first and last lines in a fixed pattern so viewers recognise the show. Three to five episodes of the same duration make a good first test of whether the look holds up.
How do you batch a season?
Write eight scripts of 20 to 45 seconds. Submit each as its own job with its own Idempotency-Key that includes the episode number, so a retry reuses the job instead of creating a second one. Use mode: "webhook" with a public HTTPS webhook_url to be told when each finishes instead of polling.
Keep the scene text identical across episodes and change only the script. Variation in the scene prompt is what makes a season look like eight different videos.
Read quality before you start. Avatar Video defaults to plus, standard is the fastest path and max is the highest tier with slower turnaround, so run one episode at each level and pick the cheapest one that looks right on a phone.
for ep in 1 2 3; do
curl -sS -X POST https://api.sume.com/v1/avatar-1.0/talking-video \
-H "Authorization: Bearer $SUME_API_KEY" \
-H "Content-Type: application/json" \
-H "Idempotency-Key: season-1-episode-$ep" \
-d "{
\"avatar_handle\": \"sume_clawra\",
\"aspect_ratio\": \"9:16\",
\"quality\": \"plus\",
\"script\": \"Episode $ep: one fact about volcanoes, in thirty seconds.\"
}"
echo
doneWhat stays yours to do?
Sume does not create the YouTube show, set serial or non-serial, upload banner and backdrop art or schedule publishing. Banner and backdrop art is a still, which you can make with the Image API at 16:9, then check against the sizes on YouTube's page. Check disclosure rules for synthetic or altered content in your channel settings before you publish an AI presenter.
Quality is also on you. YouTube's bar of high-quality video and sound applies to the whole show, so watch every episode before publishing and drop any with warped hands or off-sync speech. For a recap episode that reuses an earlier one, see a recap from one episode.
Avatar Video is limited to one avatar per video, so a two-person scene means two videos joined later, which the two-host podcast post walks through.
Sources
Related posts
More in Use cases
- YouTube Shorts 3 minutes: Oct 15, 2024 vs Dec 8, 2025 dates
YouTube's help page gives two dates for three-minute Shorts: Oct 15, 2024 for standard channels, Dec 8, 2025 for Artist Channels. How to check a file.
- YouTube Studio AI clip suggestions: 10 countries; clips elsewhere
YouTube's AI clip suggestions are limited to ten countries and select languages. Outside them, find moments from a transcript and cut podcast clips with Sume.
- YouTube thumbnail test: 3 variants, 2 weeks, desktop only
YouTube's A/B test takes up to three thumbnails and picks by watch time. Generate three distinct 16:9 candidates in one Sume image call and keep them testable.
- YouTube video A/B testing of three cuts: build them with Sume
YouTube says Studio will test up to three cuts of a video. Prepare three openings with video trim and a Timeline plan before you pay for any render.
Written by Sume