01
frame content
What we see
Subject, place, light, material and one clear action. Add precise text and logos in post.
Veo 3.1 inside Verstak
Describe what viewers see. And what they hear.
Veo designs a short scene with motion, atmosphere and native sound together.
The agent turns your idea into visual, camera and sound direction, selects a Veo tier, shows the estimate and carries the result into editing, captions and publishing.
Veo 3.1 LiteVeo 3.1 FastVeo 3.1 Quality
Veo 3.1 · 9:16 · 8snative audio
Film a bottle on wet basalt in a glass greenhouse. Macro becomes a slow orbit; we hear rain, one drop and a glass chime.
01Scenelight · material · action
02Cameramacro → slow orbit
03Soundrain · drop · chime
00:08
raindropwindchime
Verstak agent
I’ll build an 8-second vertical scene with a camera path and four timed sound events. We can start with Lite; you will see cost before launch.
Three layers of a useful brief
Veo hears a scene when sound is directed as carefully as the frame
You do not need a magic prompt. You need observable choices: what happens, how it is filmed and which sounds belong to which moment.
02camera direction
How it is filmed
One camera move, shot size, depth of field and a clear landing moment.
03sound score
♪
What we hear
Atmosphere, timed foley, speech or silence. Link every key sound to visible action.
Real Veo outputs
Two generations. Four synchronized beats.
The product scene tests material, camera and foley; the miniature story tests action, atmosphere and a sound cue. Originals and provenance are stored with the assets.
Veo 3.1 Lite
01 · macro + rain
Open with texture
Condensation and close rain give the frame physical presence before the camera moves.
Veo 3.1 Lite
02 · orbit + chime
Land motion on a cue
The orbit opens the space while one glass chime marks the final light flare.
Veo 3.1 Lite
03 · action + atmosphere
Stage a tiny story
The shutter, distant bell and birds establish place before coffee appears.
Veo 3.1 Lite
04 · foley + ending
End on an audible gesture
Steam and the cup touching wood turn the final object into a story beat.
Two tested workflows
A product and a story need different sound logic
Both demos use one visual-and-sound brief and were generated with Veo 3.1 Lite.
Veo 3.1 Lite · 8s
Product scene
Object → material → spatial reveal
Foley reinforces texture and one final visual cue.
01
macro texture
02
camera move
03
synchronized foley
00:08
Veo 3.1 Lite · 8s
Short story
Place → action → result
Atmosphere builds the world and one audible gesture closes the beat.
01
character action
02
location bed
03
final cue
00:08
Native sound is part of the generation, not a master track. For exact words, a brand voice or licensed music, the agent builds separate controlled stages.
Three working roles
Lite, Fast and Quality are an iteration strategy
The expensive tier is not for every take. The agent first helps find the scene that works.
Veo 3.1 Lite
draftThe economical way to test composition, motion and sound. 4/6/8 seconds and up to three images.
Ideas, quick variants and high-volume social scenes.
Veo 3.1 Fast
working takeA fast cinematic tier with the same visible controls and native sound.
A selected setup with several strong variants.
Veo 3.1 Quality
finalThe expensive tier for an approved scene, launched only after estimate and confirmation.
A hero take after draft selection.
Two ways to work
Delegate the scene or run it manually
Both paths share one workspace, balance, gallery and version history.
Agent: the whole scene
Describe the idea normally. The agent clarifies format, camera, sound and tier, then shows a plan.
- ✓Turns the idea into visual and audio direction
- ✓Estimates cost before launch
- ✓Continues into editing and publishing
Workshop: direct launch
Choose Veo, 4/6/8 seconds, aspect, seed and up to three images.
- ✓Lite, Fast or Quality without a hidden swap
- ✓9:16, 16:9 or 1:1
- ✓Every output becomes a new version
After generation
A clip with sound is not yet a finished format
Verstak preserves the chosen take and adapts it for the destination.
Estimate first
Tier, duration and tokens are visible before an expensive step.
One balance
Video, images, voice and music use the same balance.
Versions preserved
An accepted take does not disappear when you try another.
Full cycle
Editing, exact voice, music, captions, Reel and publishing pack.
Honest limits
Native sound is striking, but it needs review
Watch and listen to important scenes in full before approval.
Speech
Words, articulation and language can drift; use a separate voiceover for exact copy.
Synchronization
A sound event can arrive before or after the visible action.
Identity
References guide the look but do not guarantee every facial or geometric detail.
Text
Add small copy and logos after generation.
Controls
Motion transfer, source edit, camera fixed, exact first/last frames and extend are not currently promised.
Veo 3.1 questions
01What is Veo 3.1?＋
It is Google DeepMind’s video model family. In Verstak, Veo is one tool used by the content agent and the manual workshop.
02Can I use Veo without writing English prompts?＋
Yes. Describe the scene and revisions in your language; the agent prepares the production brief.
03Does Veo generate sound?＋
Yes. Available Verstak tiers return native scene sound. Use separate controlled stages for exact speech, brand voice and music.
04Can I create video from a photo?＋
Yes. The current interface accepts up to three images to guide the scene look.
05How long is a clip?＋
The current Verstak contract supports 4, 6 or 8 seconds. A longer Reel uses several clips and an edit.
06How do Lite, Fast and Quality differ?＋
Lite is economical for drafts, Fast for working variants and Quality for an approved final scene.
07How much does Veo cost?＋
Cost depends on tier and duration. The agent shows a token estimate before launch.
08Can I set first and last frames?＋
Google Veo is broader than our current UI; exact first/last frames are not yet part of the public Verstak contract.
09Can it become a finished Reel?＋
Yes. The agent assembles Veo clips, controlled sound, music, captions and a publishing pack.
10Do I need prompt skills?＋
No. Describe the scene, desired motion and key sounds; the agent builds the brief and constraints.
Describe the scene. Let the agent direct frame and sound.
See the Veo tier, duration, inputs and cost before launch.
Early access · costly steps require confirmation