AI spokesperson video for TikTok ads: a spec check against Sume
TikTok recommends 9:16 with at least 540x960 for non-Spark ads. Sume avatar videos default to 9:16 at 720p. What to verify before you upload.

For TikTok ads, TikTok recommends 9:16 with a minimum of 540x960 for non-Spark ads. Sume Avatar 1.0 renders 9:16 by default at 720p and runs 4 to 60 seconds, so its output type fits the ad spec on aspect ratio and length. Confirm the file size and bitrate on your actual export before you upload.
The TikTok side
From TikTok's video ad specifications, read 2026-10-08, for non-Spark ads:
| Item | TikTok non-Spark ad spec |
|---|---|
| Recommended ratio | 9:16, minimum 540x960 |
| Other ratios | 16:9 minimum 960x540; 1:1 minimum 640x640 |
| Formats | mp4, mov, mpeg, 3gp, avi |
| Length | Up to 10 minutes |
| File size | 500 MB or less |
| Bitrate | 516 kbps or more |
The Sume side
From Generate avatar video, read 2026-10-08:
| Item | Sume Avatar 1.0 |
|---|---|
| Aspect ratio | 1:1, 3:4, 9:16, 4:3, 16:9; default 9:16 |
| Resolution | 720p at this time |
| Duration | 4-60 seconds per job |
| Captions | Optional inline captions burned into the MP4 |
Gaps to check yourself
The Sume page does not state the container, bitrate or file size of the returned MP4, and it gives the resolution as 720p without a pixel size. Before a campaign, render one clip, run Video inspect on it, and compare the probe to the TikTok numbers above. Do not assume a clip meets the bitrate floor because it looks sharp.
A spokesperson ad is also an advertising claim made by a synthetic person. Check TikTok's own policies on AI-generated content before launch; this post covers only the file specifications.
A short workflow
- Create or reuse an avatar handle.
- Write a script of 15-30 seconds for a typical hook-and-offer ad.
- Preview first frames at 9:16.
- Render with captions enabled.
- Probe the output, then upload.
Checklist before upload
Probe the rendered file for container, resolution, duration and bitrate. Compare to the table above. Confirm captions are readable at phone size and inside the platform's safe areas. Listen to the audio on speakers and headphones. Confirm that the speaker and product claims match your policies, since an AI spokesperson still carries an advertiser's responsibility for what is said.
If a spec is missed, re-render with the same avatar handle rather than altering the file by hand. For Spark Ads, TikTok's page says mp4 or mov with no duration restriction, so the length rule above applies to non-Spark ads only.
Sources
Related posts
More in Use cases
- Allergen line on a holiday menu video: burn it in with caption cues
Add a fixed allergen and price line to a silent holiday menu clip with video-captions cues. No speech-to-text runs, and Sume does not check the allergen text.
- Amazon holiday video review: upload dates for Black Friday 2026
Amazon says holiday shoppable video review can take seven business days. Upload by Nov 17 for Black Friday 2026 and Nov 18 for Cyber Monday; clip cost inside.
- Amazon video: key benefit in the first 10 seconds, checked in 5 stills
Amazon says the best shoppable videos show the key benefit in 5 to 10 seconds. Pull five stills at 1, 3, 5, 7 and 9 seconds with Sume video-frames to check.
- Ambassador contract ended: recast six 12-second ads for $27.00 on Sume
Replace a former spokesperson in six 12-second ads with H3 Max Recast on Sume: $27.00 at 768p, $40.50 at 1080p, with the rights checks that come first.
Written by Sume