Skip to main content
Video modelsGoogleAvailable in Neiron

Veo 3.1 — video scenes with audio

With Veo, describe the scene together with sound, such as rain, street noise, or a short spoken line. Specify visual action and the audio setting separately. Review the full clip afterward: an audio track does not guarantee the requested wording or synchronization.

Content updated:

Ready-to-use prompts for Veo 3.1

Select this model in the bot or on the website. Prepare the listed material, copy a prompt and paste it into the selected model. Replace example details with your own and review the output using the checks below.

These are examples to try, not test results. Copying does not submit a task; check access and allowances in your account.

An atmospheric shot without a complex story

Before you start

No photo is needed. Describe one shot and select duration and format in the panel.

Prompt

Morning by an open window: a cup of tea sits on the table, thin steam rises slowly, and a light curtain moves gently. The camera stays still. Soft wind, no speech, music, lettering or scene changes.

Review the output

Review the steam and curtain motion. If the output contains audio, listen separately; the image does not validate the sound.

What to change in the next attempt

If unwanted sounds appear, remove the audio instructions and establish the desired visuals first.

Animate a landscape photograph

Before you start

Add one sharp landscape photo with trees or water.

Prompt

Preserve the photographed landscape. Let only leaves and the water surface move gently; buildings and the shore stay still. Keep the camera fixed. No new objects, people or sudden weather changes.

Review the output

Check the horizon, building outlines and whether extra objects appear midway through the clip.

What to change in the next attempt

If the background changes structure, request just one movement: either water ripples or moving leaves.

Compare settings and alternatives

Capabilities

Create a short scene from a text brief.

Describe the audio setting in the prompt.

Use an image as visual guidance for the scene.

Best for

Scene with ambient soundShort spoken character lineSingle-scene video announcement

Reviewed facts

Developer
Google
Purpose
Video generation and editing
Input
Text, Images
Output
Video
Developer geography
United States
Verified
2026-09-05

Benchmarks

No comparable benchmark is recorded in this card.

Sources

How to launch

  1. 1

    Open the video generator and choose the model.

  2. 2

    Add text, photos, or references depending on the scenario.

  3. 3

    Choose frame format, duration, and start generation.

Settings

Formats
16:9, 9:16
Duration
8s
Reference images
up to 1 images
Choose Veo when realistic motion and audio scene matter.
This integration currently has a fixed duration, shown below. Other Veo API modes are not promised by this card.
Check spoken wording, pronunciation, and alignment with the action. Do not treat the result as a guaranteed rendering of mandatory copy.

News examples

Compare with

Update history

FAQ

How should I prompt a video with sound?

Separate the visual brief from audio instructions. Specify whether you need music, ambience, or dialogue and which sounds to omit.

Can I rely on an exact spoken line?

You can specify dialogue, but listen to the generated track. Mandatory wording needs separate verification and, when necessary, voice production.

Does audio make Veo the best choice for every clip?

No. For a visual-only clip, compare other models’ inputs, duration, and frame format. Audio is one selection criterion.