How long can a Veo 3.1 video get? 148 seconds at 720p
Veo 3.1 extends clips 7 seconds at a time, up to 148 seconds at 720p. Sume has no extend task, so here are the longer single-clip models and how to join clips.

A Veo 3.1 video can reach 148 seconds through extension, at 720p only. Google's Veo page says Veo 3.1 and Veo 3.1 Fast extend a Veo-made video by 7 seconds, up to 20 times, with input videos up to 141 seconds and a combined output of up to 148 seconds. Veo 3.1 Lite cannot extend. Sume's Video Router does not expose an extend task, so on Sume you get length from models with longer single clips, or by joining clips in a timeline.
The numbers fit together: an 8-second first clip plus 20 extensions of 7 seconds is 148 seconds.
What are Veo's extension rules?
Most of the limits are about where the input came from. The Gemini API only extends videos that Veo generated, and the video must come from a previous generation or have been referenced in the last two days.
| Rule | Value |
|---|---|
| Added per extension | 7 seconds |
| Maximum extensions | 20 |
| Input video length | 141 seconds or less |
| Output (input plus extension) | Up to 148 seconds |
| Resolution | 720p only |
| Aspect ratio | 9:16 or 16:9 |
| Models | Veo 3.1 and Veo 3.1 Fast, not Lite |
| Input source | Videos generated by Veo only |
| Storage | 2 days; referencing a video for extension resets the timer |
What does extension do to the audio?
Google's prompting notes say extension finalizes the final second, or 24 frames, of the video and continues the action. Voice is the weak point: the page says voice cannot be effectively extended if it is not present in the last second of video. So a talking clip that ends on a pause may not carry the speaker into the extension.
Each extension also costs a generation. At Google's price list those are per-second rates, so a long chain is the sum of every segment, not one price.
What can I get on Sume instead?
Sume's Video Router catalog lists per-model limits: seedance-2.5 accepts 4 to 30 seconds at 480p, 720p or 1080p, wan-3.0 accepts 2 to 30 seconds. gemini-omni-flash-1.1 is 3 to 10 seconds, and the Video Router doc says extend and previous_interaction_id are not exposed because they are not on the provider's published input schema. Every other catalog model is capped at 15 seconds.
| Sume model id | Clip length | Resolutions |
|---|---|---|
| seedance-2.5 | 4 to 30 s | 480p, 720p, 1080p |
| wan-3.0 | 2 to 30 s | 480p, 720p, 1080p |
| gemini-omni-flash-1.1 | 3 to 10 s | 360p, 720p, 1080p, 4K |
| kling-3 | 4 to 15 s | 720p, 1080p |
How do I build something longer than one clip?
Join clips with Timeline 1.0: one audio spine up to 1800 seconds and up to 200 ordered video slots, each with start, duration and an optional transition from fade, wipe or slide types. Set the output width, height and fps once, and each slot is fitted with its fit mode (cover, contain, stretch or blur). POST /v1/timeline-1.0/plan is an unbilled preflight that returns the duration and segment count, so you can check a long assembly before you pay for it.
The trade-off against Veo extension is continuity. Extension conditions each segment on the last second of the one before; independent clips do not. Reduce the seams by ending each clip on a frame that you reuse as the next clip's first frame, and by keeping prompts consistent.
Should I plan around Veo extension at all?
Not for new work. Google's deprecations page lists the Veo 3.1 preview ids with a shutdown date of October 22, 2026 and names gemini-omni-1.1-flash as the recommended replacement, and Google's release notes describe extension for Omni as appending to the end of a clip. If your plan needs 148 seconds, decide before that date whether the unit is one long chain or a Timeline of independent clips, and test the second on a short sample first.
Sources
Related posts
More in Models
- 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.
- 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.
Written by Sume