Which Sume video models clear Instagram's 720-pixel Reel minimum
Instagram lists a 720-pixel minimum for Reels. The Sume catalog shows which video models offer 720p or above, their clip lengths, and which fall short.
Instagram's help page lists a minimum resolution of 720 pixels for Reels. When you generate a clip for one, the resolution setting is the first thing to get right. Sume's video catalog lists the resolutions each model offers, and 480p and 360p options fall short of that minimum.
The requirement is from Reel size & aspect ratios on Instagram (read 2026-10-03). It names 720 pixels without saying which side. The Sume docs do not spell out the pixel size behind each resolution label either, so treat the table as a first filter and probe the file you get back.
What each model offers
The data below is from the Sume videos docs, which describe the limits as non-uniform and send you to GET /v1/catalog for the live values.
| Model | Resolutions | Clip length | Fails the minimum at |
|---|---|---|---|
| seedance-2.5 | 480p, 720p, 1080p | 4 to 30 s | 480p |
| wan-3.0 | 480p, 720p | 2 to 30 s | 480p |
| minimax-h3 | 480p, 768p | 5 to 15 s | 480p |
| minimax-h3-max | 480p, 768p, 1080p | 5 to 15 s | 480p |
| gemini-omni-flash-1.1 | 360p, 720p, 1080p, 4K | 3 to 10 s | 360p |
| h3-max-recast | 768p, 1080p | 5 to 30 s | None listed |
Notes on the table
- The docs say 768p on the MiniMax models is a native, first-class setting and not 720p. 768 is higher than 720, but probe the output to be sure.
- Gemini Omni Flash 1.1 takes 16:9 or 9:16 only, so it suits a vertical Reel directly.
- A 480p draft is useful for testing a prompt, but a file that small would fail the minimum, so regenerate it before publishing.
Ask for the resolution explicitly
A generation request to POST /v1/videos takes model, prompt, duration, resolution and aspect_ratio. Send an Idempotency-Key so a retry is safe. This one asks for a vertical 720p clip, the lowest rung that clears the minimum on the table.
curl -X POST https://api.sume.com/v1/videos \
-H "Authorization: Bearer $SUME_API_KEY" \
-H "Content-Type: application/json" \
-H "Idempotency-Key: reel-720-001" \
-d '{
"model": "seedance-2.5",
"prompt": "A ceramic mug on a wooden desk, steam rising, soft window light",
"duration": 6,
"resolution": "720p",
"aspect_ratio": "9:16"
}'
Draft low, final high
A useful workflow is to test a prompt at the lowest resolution, then regenerate the winner at 720p or above. The catalog lists the resolutions for each model, so check it first and send a value the model accepts.
Sources
Related posts
More in Models
- Why sume/auto never picks H3 Max Recast, and what to send
A person swap needs a source video and photos. Sume's Auto router never routes to Recast, so you name h3-max-recast yourself. Here is the request.
- AI video releases and shutdowns in 2026: one dated timeline
Twelve dated events from Google, Luma, Runway, OpenAI and ElevenLabs, January to October 2026, each taken from a vendor page read on 2026-10-03.
- eleven_v4 vs eleven_v4_turbo: model IDs, endpoints, which to pick
ElevenLabs lists eleven_v4 for expressive speech with cloning in 90+ languages and eleven_v4_turbo at about 100 ms median latency. Which fits a video pipeline.
- ElevenLabs languages: Flash v2.5 has 32, Multilingual v2 29, v4 90+
ElevenLabs lists 32 languages for Flash v2.5, 29 for Multilingual v2 and 90+ for v4 and v4 Turbo. Check your markets against the model, then log it per job.
Written by Sume