YouTube suggests 3 Shorts thumbnail frames: pick from 24 yourself
YouTube suggests three frames for a Shorts thumbnail. Pull up to 24 candidates with Sume video frames and choose your own.
YouTube picks three suggested thumbnail frames for a Short on desktop, and you can pick another one on mobile. If none of those is right, extract up to 24 frames from your own clip with Sume video frames at the times you name, choose one, and upload it as the custom thumbnail. A call costs only the worker compute it uses, with no flat per-frame price.
What YouTube offers
On July 24, 2026, YouTube said YouTube Partner Program creators can upload fully custom thumbnails for Shorts, that desktop offers three suggested frames, and that mobile lets you pick any frame (YouTube Blog). The thumbnail help page lists 9:16 for Shorts, JPG or PNG files, and says Shorts thumbnails are edited in desktop Studio (YouTube Help). The blog and the help page differ on mobile, so check what your Studio shows. For a Shorts series, YouTube also lets you upload a distinct thumbnail for each Short (YouTube Help).
Pull candidate frames
POST /v1/video-frames takes one Sume-hosted clip and either an explicit at list (1 to 24 times) or an fps of at most 2, which is expanded to mid-bin samples and capped at 24. The source may be up to 300 seconds, which covers a 180-second Short. Frames keep the source size unless you set max_edge between 16 and 2160, and format is jpeg by default or png.
curl -X POST https://api.sume.com/v1/video-frames \
-H "Authorization: Bearer $SUME_API_KEY" \
-H "Content-Type: application/json" \
-H "Idempotency-Key: ep4-thumb-candidates" \
-d '{
"video_url": "https://media.sume.com/artifacts/artf_ep4/ep4.mp4",
"at": [1, 3, 5, 8, 12, 16, 20, 25],
"format": "png"
}'Read the result
The submit returns 202. Poll the video_frames resource until resource_status is ready, then read frames[{t,url,width,height}]. A frame that fails shows url: null and does not fail the whole job.
Compare what each tool gives you.
| Option | Frames offered | Control | Source |
|---|---|---|---|
| YouTube desktop suggestions | 3 | YouTube chooses the times | YouTube Blog, 2026-07-24 |
| YouTube mobile frame picker | Any frame | You scrub | YouTube Blog, 2026-07-24 |
Sume video frames at[] | Up to 24 per call | You name the seconds | Sume docs |
Sume video frames fps | Up to 24 per call, rate at most 2 | Even spacing | Sume docs |
Pick and upload
Ask for a PNG when you plan to add text or a crop, then export the final as JPG or PNG at 9:16. For a whole season, run one call per episode and keep the winners in one folder. A frame from the video itself is also the safest match to what the viewer will see after tapping. Custom Shorts thumbnails are limited to YPP channels for now, so a channel outside the program should still check its first frame; the post on pulling the frame for each episode covers the series workflow.
Sources
Related posts
More in Media tools
- YouTube thumbnails take JPG or PNG: output formats per Sume image row
YouTube accepts JPG and PNG thumbnails. Sume rows list png, jpeg or webp; Recraft V4 is webp only, Soul and Ideogram 4.5 pick the format for you. Table and fix.
- 1:1 clip on a 540x960 TikTok canvas: cover, contain or blur
A 1080x1080 clip on a 540x960 Timeline canvas: contain leaves 210 px bars top and bottom, cover cuts 420 px, blur fills the bars. All cost $0.10 a minute.
- Crop a 21:9 video to 9:16, 4:5, 1:1 and 16:9: fraction table
From a 21:9 source, a full-height crop is 0.2411 wide for 9:16, 0.3429 for 4:5, 0.4286 for 1:1, 0.7619 for 16:9. $0.02 per crop; the check is free.
- 45-minute recording: audio detach fails past 1800 s, what to do
Sume audio detach accepts a source up to 1800 seconds and returns up to 900. A 45-minute video is 2700 s, so trim first. Here is the arithmetic.
Written by Sume