Editing AI-generated video means assembling short clips into one timed cut. Every model in a production catalog returns a few seconds per generation, so a 15 or 30 second ad is four to six clips joined on a timeline, trimmed, captioned and scored. The edit decides pacing, sound and aspect ratio. The prompt never does.
You wrote a good prompt. The model returned eight seconds of a bag turning on a plinth, and it looks right. Then you open the placement brief. The ad is 30 seconds, it has to read with the sound off, and it ships in four shapes.
That gap is not a prompting problem. It is the step nobody writes about: the assembly. This guide covers what the clip length limits actually are across the models, what a finished ad needs on top of raw generations, and the order that stops you regenerating work the timeline could have fixed.
Key Takeaways
- No model gives you a finished ad. The longest single generation in the DesignerBox catalog is 20 seconds. Most sit between 4 and 10.
- A 30-second spot is four to six generations. Budget the clip count before you budget the credits.
- Captions are the highest-return edit. Meta lists them as recommended on its own Feed spec, and they cost nothing to add on the timeline.
- Fix it in the edit before you regenerate. Trimming a soft ending is free. Re-rolling the clip is not.
- One master cut, four exports. 16:9, 9:16, 1:1 and 4:5 come out of the same timeline, not four separate prompt sessions.
- The widely quoted “85% watch without sound” figure is not a Meta statistic. The sourced numbers are lower and older. Use them anyway, the direction holds.
Why every AI clip arrives too short
Generation length is capped per model, and the caps are tighter than most briefs assume. These are the durations each model page publishes on DesignerBox (designerbox.ai/models, August 2026).
| Model | Clip durations offered | Aspect ratios |
|---|---|---|
| Veo 3.1 | 4s, 6s, 8s | 16:9, 9:16 |
| Sora 2 Pro | 4s, 8s, 12s, 16s, 20s | 16:9, 9:16 |
| Seedance 2.0 | 4s, 8s, 12s, 15s | 16:9, 4:3, 1:1, 3:4, 9:16, 21:9 |
| Kling 2.6 Pro | 5s to 15s, 1-second steps | 16:9, 9:16, 1:1 |
| Runway Gen-4.5 | 5s, 10s | 16:9, 9:16, 4:3, 3:4, 1:1 |
The caps come from the models themselves, not from the plan. Google’s own documentation sets Veo 3.1 at 4, 6 or 8 seconds per generation, and requires the 8-second setting for 1080p and 4K output (ai.google.dev/gemini-api/docs/veo, August 2026).
Veo does support extension, and the limits there are worth reading before you plan around it. Google documents extending a generated video by 7 seconds at a time, up to 20 times, for a combined length of roughly 148 seconds. Extension runs at 720p only (ai.google.dev/gemini-api/docs/veo, August 2026). So the long-form route exists, and it costs you resolution.
For an ad, extension is usually the wrong tool anyway. A 30-second commercial is not one continuous shot. It is six shots.
What a 30-second ad actually needs
Take the standard shape of a direct-response spot: hook, product, benefit, proof, offer, end card. That is six beats. At 4 to 8 seconds a generation, it is six generations, and the six were never going to match on their own.
Here is what the assembly adds that no single generation gives you:
- Cuts on the beat. Raw clips run to their full length. Almost every one is stronger 1 to 2 seconds shorter.
- Order. The shot you generated third is often the hook.
- Captions. Text that carries the claim when the sound is off.
- Sound. A bed, a voiceover, or both. Most image-to-video output has no usable audio for an ad.
- The end card. The logo, the offer and the call to action are graphics, not generations.
- The export set. Four ratios from one cut.
None of that is a model job. All of it is a timeline job.
The five jobs the timeline does
The DesignerBox Video Editor is a multi-track timeline rather than a one-click assembler, which matters when you are matching a brief instead of chasing a look. Per its own feature page, it runs a base track plus up to five picture-in-picture lanes, to a maximum of 20 clips, with trim, split, ripple delete and snap, 11 transitions, 25 colour looks, and text overlays with preset animations (designerbox.ai/features/video-editor, August 2026).
The five jobs, in the order they pay off:
Trim. The cheapest quality gain available. Generated clips tend to soften in the last beat as the motion resolves. Cut before it does.
Sequence. Reordering costs nothing and changes the hook. Try the product reveal first before you accept the slow build.
Caption. Point the editor at a clip with clear speech and it writes timed captions, which you then style. If you would rather start from the transcript, the free subtitle generator does the same job as a standalone tool.
Score. A licensed music library with genre filters, a sound-effects library, AI-generated royalty-free tracks and voiceover support all sit in the same editor, so the audio pass does not become a second tool and a second licence to track.
Grade. Twenty-five colour looks, plus blur, vignette, speed and fade. One look applied across every clip is what makes six separate generations read as one shoot.
Output is MP4 at 720p or 1080p in 16:9, 9:16, 1:1 or 4:5, and a render typically lands in 1 to 5 minutes (designerbox.ai/features/video-editor, August 2026).
Captions, and the statistic everyone gets wrong
You have read that 85% of Facebook video is watched without sound. It is worth knowing where that number comes from, because it is not from Meta.
It traces to 2016 reporting in Digiday, which quoted individual publishers describing their own traffic. LittleThings said roughly 85% of its Facebook viewership happened without sound, and Mic reported a similar share of its 30-second views. Those are publisher figures for publisher content from a decade ago, not a platform-wide measurement of ads.
The better-sourced number is a Verizon Media consumer study of 5,616 US respondents, reported in 2019: 69% said they watch video with the sound off in public places, and 80% said they were more likely to finish a video when captions were available.
Both numbers are old. Neither is a controlled test of your ad. What holds up is the platform’s own guidance: Meta’s ads guide lists both sound and subtitles as optional but recommended on its Feed video spec (facebook.com/business/ads-guide, August 2026).
Caption anyway. It is a two-minute pass on the timeline, and it is the only edit that makes a silent viewer able to follow the claim.
Export once, ship four shapes
Meta’s published Feed video spec is 4:5 at 1440 x 1800 pixels, MP4, MOV or GIF, from 1 second to 241 minutes, up to 4 GB (facebook.com/business/ads-guide, August 2026). Stories and Reels are full-screen vertical. In-stream and YouTube are landscape. Nothing you build for one of them fits the others.
Generating each shape separately is the expensive mistake. Two 9:16 generations of the same brief will not match, because you are rolling the dice twice. Cut the master once, then export the ratios, and the shots stay identical across placements.
For a repeatable version of that step across channels, the multi-platform export workflow reformats one master while holding the focal point, which is the part a naive crop destroys.
Where the edit belongs in the production order
Put the edit before the second generation round, not after it.
The instinct is to keep prompting until the clips are perfect, then assemble. That order pays for problems the timeline solves for free. A clip that ends weakly does not need a re-roll, it needs a trim. A cut that feels slow does not need new footage, it needs a reorder. A shot that reads too cool does not need a new prompt, it needs the grade applied to all six.
The working order:
- Generate one clip per beat, cheapest usable setting.
- Assemble a rough cut and watch it once, silent.
- Trim, reorder, grade.
- List only the shots that still fail. Regenerate those.
- Caption, score, export.
Step 4 is usually one or two clips, not six. That is the whole saving, and it lands on your credit balance directly, since video is priced per second of output rather than per clip.
What this changes about how you prompt
Once you plan for an edit, you prompt differently.
Ask for one action per clip, not three. A clip that tries to pan, reveal and cut inside six seconds gives the timeline nothing to work with. Six clean single-action clips cut into a better ad than three busy ones.
Leave handles. Generate slightly longer than the beat needs so you have material to trim into.
Stop asking the model for text. On-screen type belongs on the timeline, where you can edit a price without paying for a new generation, and where it comes out sharp.
Match the source. Shots derived from the same product photo cut together far more easily than shots built from separate text prompts, which is the same reason one product photo can carry a whole video ad.
If you are batching this for paid social, the arithmetic of how many clips a campaign needs is worked through in the guide to AI video ad packs, and the platform-specific version for short-form sits in how to create TikTok video ads with AI.
Start from the clips you already have
Generate the beats, assemble them in the Video Editor, caption the cut and export the four shapes from one timeline. If you would rather start from a brief than from a blank timeline, the Video Ad Composer builds the structure first and hands you the cut to refine.
Every model in the catalog sits behind the same subscription, so switching models per shot costs you a dropdown, not another bill.
FAQ
Can AI generate a full 30-second ad in one clip?
Not as a single generation from the models in this catalog. The longest clip length published is 20 seconds, on Sora 2 Pro, and most models sit between 4 and 10 seconds (designerbox.ai/models, August 2026). A 30-second ad is assembled from several generations on a timeline.
How many clips does a 30-second ad need?
Four to six for a standard direct-response structure: hook, product, benefit, proof, offer, end card. Generate one clip per beat, then trim each by a second or two in the edit.
Do AI video clips come with usable audio?
Some models generate native audio, but ad audio usually needs a music bed, a voiceover or both, mixed to a consistent level across shots. That is a timeline job. The Video Editor includes a licensed music library, sound effects, AI-generated royalty-free tracks and voiceover support (designerbox.ai/features/video-editor, August 2026).
Should I add captions to AI-generated video ads?
Yes. Meta lists subtitles as recommended on its Feed video spec (facebook.com/business/ads-guide, August 2026), and captions are the only edit that lets a silent viewer follow the claim. The editor writes timed captions from a clip with clear speech, then you style them.
Can I export one edit to every placement size?
Yes. The Video Editor exports MP4 at 720p or 1080p in 16:9, 9:16, 1:1 and 4:5 from the same timeline (designerbox.ai/features/video-editor, August 2026). Cut once, export four times, and the shots stay identical across placements.
Is it cheaper to fix a clip in the edit or regenerate it?
Fix it in the edit where you can. Trimming, reordering and grading cost nothing. Regeneration is priced per second of output, so re-rolling six clips to solve a pacing problem spends credits on something the timeline already handles.
What resolution should I export ad video at?
1080p covers every placement in practice. Meta’s Feed spec recommends 1440 x 1800 for 4:5 (facebook.com/business/ads-guide, August 2026), and platforms downscale rather than reject. Note that Veo’s extension feature runs at 720p only, so long-form built by extension will not reach 1080p (ai.google.dev/gemini-api/docs/veo, August 2026).
Sources
- Veo 3.1 clip durations, resolution requirements and extension limits: ai.google.dev/gemini-api/docs/veo, accessed August 2026
- Model clip durations and aspect ratios: designerbox.ai/models, accessed August 2026
- Video Editor capabilities, export formats and render times: designerbox.ai/features/video-editor, accessed August 2026
- Meta Feed video ad specification and caption guidance: facebook.com/business/ads-guide, accessed August 2026
- Sound-off viewing figures: Digiday reporting of publisher-supplied data, 2016; Verizon Media consumer study of 5,616 US respondents, reported 2019
Model clip durations and editor capabilities verified from provider documentation and product pages as of August 2026. Sound-off viewing figures are third-party survey data from 2016 and 2019 and are cited with those dates. Individual results vary.
