AI video generator for TikTok: which Sume models make 9:16

Six Sume video models list 9:16 and one never takes an aspect ratio. Table of ratios, seconds and resolutions, read from the catalog on 2026-10-07.

4 min readSume
All posts

TikTok is a vertical format, so the first filter on any video model is whether it can output 9:16. On Sume, send aspect_ratio on POST /v1/videos and check the catalog with GET /v1/videos/models, which lists supported_aspect_ratios per model.

Vertical support by model

Limits are from the Sume catalog docs and the router constraints on 2026-10-07.

Vertical output by model (read 2026-10-07)
ModelSecondsResolutions9:16
seedance-2.54-30480p / 720p / 1080pyes
seedance-2, 2-fast, 2-mini4-15480p / 720p / 1080pyes
wan-3.02-30480p / 720p / 1080pyes
kling-34-15720p / 1080pyes
gemini-omni-flash-1.1 (sume/auto)3-10360p / 720p / 1080p / 4Kyes (16:9 or 9:16 only)
grok-imagine-video-1.54-15480p / 720pno aspect_ratio field; image-to-video only

Two cases to handle

Grok takes a still and no aspect_ratio, so the frame you send is the thing you control. For a TikTok-shaped result send a vertical image. The MiniMax H3 models take 5 to 15 seconds; check supported_aspect_ratios in the catalog before relying on vertical output, as I did not verify it.

Length

Sociality.io reports TikTok in-app recording up to 10 minutes and uploads up to 60 minutes. Any single AI job on Sume is at most 30 seconds, so a full-length video is built from several jobs and joined.

Sources

Related posts

More in Models

All Models posts

Written by Sume