YouTube ad limits for profanity on a Short: title, thumbnail, audio

Profanity limits ads on a Short in three tiers: obscured or mild words earn full ads, moderate profanity in the title or thumbnail or used in most sentences earns limited ads, and stronger profanity in a title or thumbnail, or hateful slurs anywhere, earns no ads. The page says these policies apply to video, Short or live stream, thumbnail, title, description and tags.

This comes from Advertiser-friendly content guidelines, read on 2026-10-03. It is a summary for producers, and YouTube's own page governs.

The three tiers for inappropriate language

The inappropriate language section sorts by how strong the word is, how often it appears, and where it sits.

The opening of the page says profanity or vulgarity at the start or throughout the majority of a video may not be suitable for advertising. I did not find a stated number of seconds for language, unlike other sections of the page. So treat the opening seconds as the place to be careful, not as a measured threshold.

What this means for an AI-voiced Short

A script written or generated for a voiceover can carry stronger words than you intended, and a bleeped or obscured version sits in a different tier from the spoken one. The page lists abbreviated or obscured profanity under full ads, though it does not say a bleep guarantees that outcome. The title and thumbnail are stricter than the video itself, because the same word that is fine in speech can limit ads in a title. Check all three separately.

Check the first seconds of a Sume clip

Video inspect reads one clip on media.sume.com and, with transcribe: true, returns a transcript with words. Search the transcript text for words you do not want, and look at the opening words with their timestamps. Video frames pulls stills at the seconds you pick, so you can read on-screen text in the opening shots.

If a word sits in the opening, edit the script and regenerate, or cut around it with Video trim (start plus end or duration, with audio as keep or drop).

curl -X POST https://api.sume.com/v1/video-inspect \
  -H "Authorization: Bearer $SUME_API_KEY" \
  -H "Content-Type: application/json" \
  -H "Idempotency-Key: ad-check-001" \
  -d '{
    "video_url": "https://media.sume.com/artifacts/artf_demo/short.mp4",
    "transcribe": true,
    "frames": { "at": [0, 1, 2, 3] }
  }'

Captions are part of the check

Burned-in text counts as part of the video a viewer sees. Video captions can burn the transcript or exact script_text, or authored cues for a silent clip. It costs $0.20 per job for videos up to 60 seconds under the docs' current fixed estimate, and live pricing is in GET /v1/catalog. Decide the wording you want on screen before you burn it; the related captions and profanity post covers how YouTube's own captions treat flagged words.

Sume does not predict YouTube's ad suitability decision or monetization status. It helps you read your own clip so you decide what to change.

Quick list before upload

Run this once per Short:

• Search the transcript for strong words, and check the first words spoken.
• Check the title and thumbnail text against the tiers above, since they are stricter than speech.
• Make sure no slur appears in the video, thumbnail or title.
• If monetization matters, read the page again on the day you publish, because it can change.