How to write prompts for Veo 3.1
Veo 3.1 turns a sentence or a start picture into a short clip, with sound on request. The first two sections put Google’s own prompt guide in our words, and we did not test them. The last rests on two sample clips made with this model that we watched ourselves, a small sample.
Written from the provider’s documentation and from results we checked ourselves. Last checked November 8, 2026.
What to put in a Veo prompt
From the provider’s documentation
- Google’s guide breaks a prompt into parts and says you need not use all of them: the subject, the action, the scene, the camera angle and movement, lens effects, the style and light, and the pacing.
- Be specific about the subject, naming a profession, a breed or a creature rather than a generic word. Describe the action as movement, interaction and feeling, and put actions in order if you want a story.
- For the scene, give the place, the time of day, the weather and the atmosphere, such as dust or haze.
- For short clips, the guide says to keep changes small and gradual, like a bud slowly opening or a candle burning down a little. It notes that events which take a long time do not fit in a short clip.
Camera, sound and exclusions
From the provider’s documentation
- Name the viewpoint (eye level, low, high, top-down, close-up, wide, over-the-shoulder, first-person) and the move (static, pan, tilt, dolly, truck, pedestal, zoom, crane, drone, handheld, whip pan, arc). The guide stresses that a dolly moves the camera while a zoom only changes the lens.
- It warns that some advanced camera angles and lenses are not officially supported, so results may vary.
- For sound, say plainly that you want it and write the sound in sentences of its own: a distinct effect tied to something in the scene, the ambient noise of the place, or dialogue with the speaker named before the line.
- To leave something out, list it as a plain noun, such as wall or frame, instead of writing no or don’t. Google’s example shows that a list of exclusions changed the result of an otherwise identical prompt.
- Google’s model page lists English as the prompt language.
What our sample clips showed
From results we checked ourselves
- The fisherman prompt opens with the kind of shot, “Documentary shot: …”, then who, where, small details and the light. The temple prompt opens with the scene and ends with a camera move.
- The fisherman clip shows an old man mending a net with gulls overhead and golden light. Part-way it cuts to a wide side view of him on the pier, although the prompt named no cut. We could not make out the visible breath in the cold air that the prompt asked for.
- The temple clip shows rain, ripples in the stone basin, glistening moss and a camera that drifts closer. The bamboo spout pours water into the basin, and we did not see it tip over as the prompt asked.
- Two clips are too few to say how often this happens. We did not judge the sound.
Real results and the prompts behind them
Results made with this model. The prompts are shown exactly as written, in English.
Settings of this model
- Von
- Typ
- AI video model
- Erstellungsart
- Text zu Video · Bild zu Video · Referenz zu Video
- Länge
- 4–8s
- Auflösung
- 720p / 1080p / 4K
- Aspect ratios
- 16:9 · 9:16
- Sound
- Mit Ton
- Preis
- from 336 / sec
Fragen
How should I ask for sound with Veo 3.1?
Google’s guide says to state that you want audio and to describe it in separate sentences: a sound effect, the ambient noise or the dialogue, with the speaker named. We have not judged the sound of our own clips.
Why did my clip not show everything the prompt said?
In each of our two samples one requested detail did not show: breath in cold air could not be seen, and a bamboo spout poured instead of tipping. Google’s guide advises keeping changes small and gradual in a short clip, so ask for one clear action and check the result.
Can I use camera terms like dolly zoom?
Google says some advanced camera angles and lenses are not officially supported and results may vary. Use the terms the guide lists, such as pan, tilt, dolly or handheld, and keep a dolly, where the camera moves, apart from a zoom, where the lens changes.
Provider documentation we read
We read these pages and wrote the points that matter in our own words. The provider may change them.
- Veo prompt guide (Google Cloud) (docs.cloud.google.com)
- Veo 3.1 model page (Google Cloud) (docs.cloud.google.com)
More prompting guides
- How to write prompts for GPT Image 2.6 FlareAI image model
- How to write prompts for GPT Image 2.6 SunburstAI image model
- How to write prompts for Nano Banana 2.1AI image model
- How to write prompts for MOSS ImageAI image model
- How to write prompts for Grok Imagine ProAI image model
- How to write prompts for Nano Banana ProAI image model
- How to write prompts for Nano Banana 2AI image model
- How to write prompts for GPT Image 2.5 FlareAI image model
- How to write prompts for IndigoAI image model
- How to write prompts for SaffronAI image model
- How to write prompts for CeladonAI image model
- How to write prompts for LapisAI image model
- How to write edit requests for UmberKI-Bildeditor
- How to write prompts for Kling Video 4.0AI video model
- How to write prompts for Seedance 2.5AI video model
- How to write prompts for Seedance 2.0 MiniAI video model
- How to write prompts for Kling V3AI video model
- How to write prompts for Kling O3 OmniAI video model
- How to write prompts for Veo 3.1 FastAI video model
- How to write prompts for SilkAI video model
- How to write prompts for LacquerAI video model
- How to write prompts for SatinAI video model
Die Namen von Modellen, Unternehmen und Produkten auf dieser Website sind Marken ihrer Inhaber; Pickshot ist mit ihnen weder verbunden noch wird es von ihnen gesponsert oder unterstützt.