Veo 3.1 Lite 1080p is 8 seconds only: options for a 5-second clip
Veo 3.1 Lite renders 1080p only at 8 seconds. To get a 5-second 1080p clip, trim an 8-second render with Sume or pick a model whose range includes 5.

You cannot ask Veo 3.1 Lite for a 5-second clip at 1080p. Google's Veo page lists 4, 6 and 8 seconds for Lite and says the duration must be 8 seconds at 1080p, so a 1080p render is always 8 seconds long. There are two ways around it: render the 8 seconds and cut it to 5 with Sume's video trim, or use a model whose 1080p range already includes 5 seconds.
The second route is also the one that lasts. Google's deprecations page lists veo-3.1-lite-generate-preview with a shutdown date of October 22, 2026 and gemini-omni-1.1-flash as the recommended replacement.
What exactly does Google say about Lite durations?
The Veo page and the pricing page give a short list of Lite constraints. 4k is not available on Lite at all, and the model is still labelled Preview.
| Item | Veo 3.1 Lite |
|---|---|
| Durations | 4, 6 or 8 seconds |
| Duration at 1080p | 8 seconds only |
| Resolution | 720p (default) or 1080p; 4k not supported |
| Price per second (with audio) | $0.05 at 720p, $0.08 at 1080p |
| Status | Preview, shutdown October 22, 2026 |
Route 1: render 8 seconds, then trim to 5
At $0.08 a second, an 8-second 1080p Lite clip costs $0.64 on Google's price list. Sume's video trim cuts a [start, end) range out of one Sume-hosted clip into a new MP4 for $0.02 per job, re-encoding frame-accurately by default and keeping the audio unless you pass audio: drop. Sume's catalog does not list Veo, so the Veo file has to be imported into your workspace first: the trim doc says there is no open-internet fetch and points to POST /v1/media-imports.
Pick the 5 seconds you want from the 8 and set start. The example below keeps seconds 1 to 6.
curl -X POST https://api.sume.com/v1/video-trim \
-H "Authorization: Bearer $SUME_API_KEY" \
-H "Content-Type: application/json" \
-H "Idempotency-Key: veo-lite-trim-001" \
-d '{
"video_url": "https://media.sume.com/artifacts/artf_demo/veo-lite-8s.mp4",
"start": 1,
"duration": 5
}'Route 2: a model whose 1080p range includes 5 seconds
Sume's Video Router catalog gives per-model limits. gemini-omni-flash-1.1 takes 3 to 10 seconds at 360p, 720p, 1080p or 4K in 16:9 or 9:16, with native synced audio. Google's release notes say 1080p and 4K Omni outputs are generated using upscaling, so treat 1080p there as an upscaled delivery size.
Sume bills the provider list price times 1.25 on every row. The list rates in the table are from the Video Router API doc.
| Sume model id | Duration | 1080p | Provider list per second |
|---|---|---|---|
| gemini-omni-flash-1.1 | 3 to 10 s | Yes (360p to 4K offered) | $0.03 / $0.10 / $0.15 / $0.30 at 360p / 720p / 1080p / 4K |
| kling-3 | 4 to 15 s | Yes (720p, 1080p) | See the catalog row |
| seedance-2.5 | 4 to 30 s | Yes (480p to 1080p) | See the catalog row |
| wan-3.0 | 2 to 30 s | Yes (480p to 1080p) | $0.05 / $0.10 / $0.20 by resolution tier |
Which route should I use?
If the clip must come from Veo's look, trimming is the only option, and you pay for 3 unused seconds. If any good model will do, ask for exactly 5 seconds and skip the cut: a 5-second 1080p Omni clip is $0.75 at the $0.15 list rate before Sume's 1.25 multiplier.
Either way, read capabilities from GET /v1/video-router/models before you hard-code a duration, because limits differ per model.
How do I check the result before publishing?
After a trim or a short generation, confirm the length and audio track instead of trusting the request. Sume's video inspect returns probe facts such as probe.has_audio, and its stills let you look at the first and last frames of a 5-second cut to be sure the trim did not clip the action.
Keep the 8-second master as well. A trim never touches its source, so if the 5-second version turns out too tight you can cut a different range from the same file for $0.02 instead of paying for another render. That matters more on Veo than on cheaper models: a 1080p Lite clip is $0.64 every time you regenerate it, while a second trim is two cents.
One more scheduling point. Because the Lite preview id is scheduled to stop on October 22, 2026, a pipeline that depends on its 8-second rule will need a new duration plan anyway. Writing the duration into one config value, and choosing it per model from the catalog, makes that migration a one-line change instead of a rewrite.
Sources
Related posts
More in Models
- Veo 3.1 Lite takes text and images only: no references or extension
Google's Veo page shows Lite without reference images, video input or extension. If your Lite job needs those, here is what to use instead on Sume.
- 9:16 vertical AI video: which Sume models take an aspect ratio
Seedance, Wan 3.0, Kling 3, MiniMax H3 and Gemini Omni Flash take 9:16; Grok Imagine, Genjutsu and H3 Max Recast take no aspect ratio. A per-model table.
- AI video ratios on Sume: Omni Flash is 16:9 or 9:16, Kling adds 1:1
Gemini Omni Flash 1.1 on Sume takes only 16:9 and 9:16; Kling 3 adds 1:1. Which other Sume video models list more ratios, and which list none.
- Video-to-video clip length limits on Sume, by tool
Recast takes 5 to 30 s, Genjutsu 4 to 30 s, Omni edit 3 to 10 s, Avatar face swap about 4 to 15 s. A table and a Python picker choose by clip length.
Written by Sume