How to Generate a Short Video

Learn how to create AI-generated videos using VEO 3.1, choose duration and resolution, and understand credit costs.

Home › Help Center › AI Features › How to Generate a Short Video

Generate stunning AI videos in seconds using our video generation models. This guide covers everything from writing your prompt to choosing the right model and resolution settings.

Watch the 1-minute walkthrough — generate a clip and publish it to TikTok

Quick Start

Navigate to Create Content or Flow — You can generate videos from the Create Content page or from within a Flow project. Both share the same model catalog and pricing. Enter your prompt — Describe the scene, action, camera movement, lighting, and style. Specific prompts produce much better results. Pick aspect ratio — Choose Landscape 16:9 or Portrait 9:16 at the top of the storyboard. Portrait is ideal for TikTok, Reels, and Shorts. Select your model and duration — See the model table below. The Generate button always shows a [Cost | X left] badge so you know the credit cost before committing. Click Generate — Your clip appears in the storyboard when ready. From there you can preview, regenerate, extend, or publish.

Available Models

All video models are available on every plan — credit cost is what scales with quality and duration.

Model	Durations	Resolutions	Audio	Best For
Simplest	4s / 6s / 8s	720p, 1080p	Native (free)	Budget-friendly drafts with audio
Fast with sound	4s / 6s / 8s	720p, 1080p, 4K	Native (free)	Balanced speed + quality, default choice
Rich motion	4s / 6s / 8s	720p, 1080p, 4K	Native (free)	Hero cinematic clips
Sharpest people	3s – 15s (every second)	720p, 1080p	Optional toggle (+50%)	Long clips, image-to-video with start/end frame
Detailed scenes	4s – 15s	480p, 720p	Native	Realistic motion + Reference-to-Video
Quick scenes	4s – 15s	480p, 720p	Native	Cheaper detailed scenes for iteration
Budget realistic	5s / 10s	720p, 1080p	None	Lowest-cost realistic everyday clips

Resolution Rules

Fast with sound & Quality

Resolution	Best For	Requirement
720p	Quick previews, stories	Available at all durations
1080p	Standard social posts	Requires 8s duration
4K	Hero clips, large displays	Requires 8s duration

Auto-downgrade: If you select 1080p or 4K with a duration shorter than 8s on Fast with sound/Quality, the resolution is automatically downgraded to 720p to keep the generation valid. Simplest has no 4K option but works at 1080p for all durations.

Sharpest people

720p and 1080p available across all durations from 3s to 15s. No duration restriction on resolution.

Detailed scenes

Caps at 720p — pick this model for motion realism rather than max resolution.

Credit Costs

Costs come straight from the live pricing matrix. Below are the most common configurations:

Model	4s	5s	6s	8s	10s
Simplest (1080p, audio)	51	—	77	102	—
Fast with sound (1080p, audio)	77	—	115	153	—
Rich motion (1080p, audio)	254	—	381	508	—
Sharpest people (1080p)	107	134	160	213	267
Detailed scenes (720p, audio)	192	240	288	384	480
Quick scenes (720p, audio)	153	191	229	305	381
Budget realistic (720p)	—	40	—	—	80

Sharpest people is billed at a flat 27 credits per second, audio included. The Simplest, Fast with sound and Rich motion tiers include native audio at no extra charge. Budget realistic does not generate audio at all.

Reference Image (Image-to-Video)

Every model in the catalog accepts a reference image as the starting frame. Upload one from your device or pick one from your Media Library.

Sharpest people additionally supports a last frame — set both a start and end image and it interpolates the motion between them.
Fast with sound / Quality / Lite support a single start frame plus last-frame interpolation.
Detailed scenes (Quality & Fast) uniquely supports Reference-to-Video (Ref2V), where multiple reference images guide the character/scene rather than acting as the first frame.
Budget realistic supports image-to-video (start frame only).

Audio

Simplest, Fast with sound and Rich motion always include native audio — ambient sound, effects, and dialogue when prompted — at no extra credit cost.

Sharpest people has an audio on/off toggle. Audio off is cheaper; audio on adds roughly 50% to the cost.

Detailed scenes generates native audio by default.

Budget realistic is video-only — there is no audio toggle.

Publishing Your Generated Video

Once a clip is generated, you can publish it directly to your connected social accounts without leaving the page. The video in this guide shows publishing to TikTok.

Click Post / Publish on the generated clip card to enter the Preview & Publish step. Select platforms — Pick TikTok (or any combination of TikTok, Instagram, YouTube Shorts, Facebook, X, LinkedIn). Each selected platform reveals its own options panel. Fill in the TikTok options — Title, privacy level (Public / Friends / Followers / Only Me — pulled live from your TikTok account), and toggles for comments, Duet, and Stitch. If your post features commercial content or branded music, tick the music usage and branded content disclosures so the post passes TikTok review. Write your caption — AI suggestions are available, and captions auto-truncate to each platform's limit (TikTok 2,200 / X 280 etc.). Publish Now to send immediately, or Schedule to pick a date and time. Scheduled posts publish automatically via the queue. TikTok requires that you complete the music-usage declaration before the Publish button enables — it's a TikTok API requirement, not an ACE limitation. For best results, write detailed prompts that describe camera movement, lighting, and mood. For example: "A slow cinematic pan across a neon-lit Tokyo street at night, rain reflecting off the pavement, cyberpunk atmosphere."