Veo 3 Prompts: 10 Tips + Copy-Paste Templates (2026)

You type cinematic, 8K, masterpiece into a box and hit generate. Eight seconds later the clip looks like a stock wallpaper that learned to breathe. The chorus still has no camera. The singer’s mouth is moving, but nothing in the room is making sound. You did not fail at “being more creative.” You handed the model a mood board and asked it to direct a shot.

People searching veo 3 prompt and prompt for veo 3 already picked the model. What they want is a string they can paste: camera first, then subject, action, room, light — plus a separate audio sentence if the clip is allowed to make noise.

This is not a methodology lesson. Every tip below is a full block. The anti-example is the same intent written as empty adjectives. Paste the block, pick Veo 3.1 Fast or Veo 3.1 Lite in the SunoMV audio-to-video generator, and export a shot that can sit on a chorus.

Table of Contents

• Why “cinematic, 8K” is not a Veo 3 prompt
• The official five slots Google actually uses
• 10 copy-paste tips
• 6 music-video templates
• Empty prompt vs a prompt for Veo 3
• FAQ
• Copy, pick Veo, export the MV

Why “cinematic, 8K” is not a Veo 3 prompt

Google’s own Veo 3.1 prompting guide puts cinematography in slot one on purpose: camera work is “the most powerful tool for conveying tone and emotion.” If you skip it, the model invents a default wide shot and a default pan, then spends the rest of the clip looking expensive.

The same guide locks two physical facts you should write as fields, not as vibes. Clips are 4, 6, or 8 seconds. Aspect ratio is 16:9 or 9:16. In SunoMV, Veo shots lock to 8 seconds — so a prompt that describes a 20-second tracking move will get chopped, not directed.

I’d rather write one boring, complete sentence per slot than a paragraph of “epic energy.” Empty adjectives are how you get a pretty clip that cannot cut to a downbeat.

Image: SunoMV Team · music-video still written as camera + subject + rain, not as “cinematic 8K”

Practical rule: If the first clause is not a camera, you are decorating a shot the model already chose for you.

The official five slots Google actually uses

Google Cloud’s formula is five parts, in this order: cinematography + subject + action + context + style and ambiance. Independent write-ups such as Prompt Architects’ Veo 3 prompt structure restated that same order after Google’s Veo 3.1 guide. The Gemini API Veo page (updated 30 July 2026) then adds an audio layer on top: dialogue in quotes, sound effects named as sounds, and ambient noise as its own sentence.

Fill the table. Any empty cell becomes the blandest default.

Slot	Fill this	Anti-example
Cinematography	Shot size + angle + named move + speed	”cinematic camera”
Subject	One person, wardrobe, age band, hair	”a beautiful girl”
Action	Verbs + physics (weight, speed, contact)	“dancing beautifully”
Context	Room, weather, time of day, foreground clutter	”epic location”
Style & ambiance	Light direction + texture + color grade	”8K, masterpiece, highly detailed”
Audio (optional)	Dialogue / SFX / ambient as separate sentences	”add dramatic music”

DeepMind’s Veo page is blunt about what Veo 3 actually sells: native sound effects, ambient noise, and dialogue, generated with the picture. If you skip the audio sentence, you still get a soundtrack — just not the one the chorus needs.

In SunoMV the picker shows Veo 3.1 Fast and Veo 3.1 Lite. Lite is the silent, faster option. If the prompt for Veo 3 includes spoken lines or named SFX, pick Fast. A silent option will ignore the audio block.

Practical rule: Write the five picture slots in Google’s order, then add audio as its own sentences. Do not bury “she whispers” inside the wardrobe line.

10 copy-paste tips

Each tip is a full block. The anti-example is the same intent as empty adjectives. In SunoMV: pick Veo 3.1 Fast or Lite, paste the whole block into the shot description, generate, then cut the 8-second clip onto the lyric line.

1. Put the camera in the first clause

One-line rule: Google put cinematography first because the model obeys named moves more than mood words. The same craft shows up in Runway’s camera-prompt notes: “slow dolly push-in” beats “cinematic.”

Medium close-up, low angle, slow dolly push-in over 8 seconds,
on a singer in a soaked black leather jacket standing in neon rain.
35mm spherical look, shallow depth of field, magenta rim light from the left.

Anti-example: Cinematic music video of a singer, highly detailed, 8K.

How to use it in SunoMV: paste this as the first transition on a chorus downbeat. The push-in is the cut, not the outfit.

2. Name one subject the way a costume department would

One-line rule: “a girl” is a casting call. Wardrobe is a lock.

Subject: a woman in her late 20s, blunt black bob, silver hoop earrings,
oversized oxblood leather jacket over a white tank, chipped black nail polish.
Keep the same jacket and hair in every shot.

Anti-example: A beautiful mysterious girl with good vibes.

How to use it in SunoMV: copy this subject block into every Veo shot for the same song so the chorus and the verse do not recast her.

3. Write action as physics, not as a compliment

One-line rule: “dancing beautifully” has no weight. “Heel strikes wet asphalt” does.

Action: she takes three slow steps toward camera, left heel striking wet asphalt,
jacket hem swinging, rainwater flicking off the collar.
She does not spin. She does not jump. Weight stays in the front foot.

Anti-example: She dances energetically with amazing choreography.

How to use it in SunoMV: use this on a pre-chorus walk-up. High-burst flips are where Veo shots fall apart; keep the body slow and let the camera move.

4. Make the room do work

One-line rule: context is not “location: cinematic.” It is the stuff the lens can hit.

Context: a narrow Tokyo side street at 1 a.m., vending machines buzzing on the right,
a puddle in the foreground reflecting pink kanji, steam from a ramen grate,
one passing taxi light in the deep background. No crowd.

Anti-example: Epic futuristic city, breathtaking scenery.

How to use it in SunoMV: one concrete street per song. Reuse it. The puddle is a free transition device.

5. Replace “cinematic lighting” with a direction and a texture

One-line rule: light direction is the cheapest quality lever in the official samples.

Style and ambiance: hard key from camera-left neon, magenta, cutting a
narrow rim on her jaw. Soft fill from a shop window on the right.
Low-key contrast, wet highlights, slight grain, no clean beauty lighting.

Anti-example: Cinematic lighting, volumetric god rays, ultra realistic.

How to use it in SunoMV: keep the same key/fill pair across shots so the grade does not reset every 8 seconds.

Image: SunoMV Team · the first clause is the camera move, not the adjective “cinematic”

6. Put dialogue in quotation marks, on its own line

One-line rule: the Gemini API audio notes tell you to mark speech with quotes. If it is not quoted, it is narration, not a line.

She looks just past camera and says, "Don't wait for the chorus."
Keep her mouth in frame. No other speakers.

Anti-example: She talks emotionally about the song.

How to use it in SunoMV: only with Veo 3.1 Fast. If you picked Lite, delete this block — a silent clip will mime the line and you will hate the result.

7. Name SFX as sounds, not as feelings

One-line rule: “dramatic impact” is a mood. “Heel click on wet brick” is a cue.

SFX: a sharp heel click on wet brick; a distant train door chime;
rain ticking on a metal awning. No explosion. No whoosh transition.

Anti-example: Add epic sound design.

How to use it in SunoMV: write SFX that can land on a snare. If the song already has a dense mix, keep SFX sparse so they do not fight the track.

8. Give the room a bed of ambient noise

One-line rule: Google lists ambient noise as a third audio type. Skip it and the model invents a generic music bed.

Ambient noise: low traffic two streets over, a vending machine hum,
rain on plastic sheeting. No score. No vocal ad-libs from off-screen.

Anti-example: Background music that fits the mood.

How to use it in SunoMV: for lyric videos you already have a song. Tell Veo no score so the native bed does not double the chorus.

9. Treat 16:9 and 9:16 as a slot, not a crop

One-line rule: Google’s Veo 3.1 guide only offers 16:9 or 9:16. A 1:1 prompt is a crop you will do later, badly.

Aspect ratio: 9:16 vertical. Headroom for captions. Subject stays in the
center third. No wide establishing shot. Phone-first framing.

Anti-example: Make it work for every platform.

How to use it in SunoMV: one ratio per export. Do not generate 16:9 and hope a vertical crop keeps the face.

Image: SunoMV Team · 9:16 as a written slot, not a crop of a wide shot

10. Write the 8-second lock instead of fighting it

One-line rule: in SunoMV, Veo shots are fixed at 8 seconds. A prompt that asks for a minute-long oner will not get a minute.

Duration: one 8-second shot. Start on her hands at the jacket zipper,
end on her eyes as she looks up into the rain.
No cut inside the clip. No time-lapse. No "then later that night."

Anti-example: A full music video that follows her all night across the city.

How to use it in SunoMV: one prompt = one lyric line or one downbeat. String the 8-second clips. Do not ask one Veo shot to be the whole MV.

Practical rule: If the prompt contains “then,” you are writing a sequence. Split it. Veo will not storyboard a paragraph for you.

6 music-video templates

Each template is one paste. Swap the bracketed bits. Keep the slot order.

Pop chorus hook (16:9)

Cinematography: medium shot, eye-level, slow push-in, 35mm.
Subject: a woman in her 20s, red vinyl jacket, wet hair stuck to her cheek.
Action: she mouths the chorus and steps one pace toward camera, rain on her lashes.
Context: rooftop at night, city grid behind her, one red aviation light blinking.
Style and ambiance: hard backlight, thin fog, teal-and-red grade, light grain.
Audio: she sings, "Stay until the lights go out." Ambient: wind on a metal rail.
No score. 16:9. 8 seconds.

Lyric close-up (mouth + caption room)

Cinematography: close-up, slight low angle, locked-off with a 2cm drift.
Subject: the same woman, oxblood jacket, silver hoops, a small scar on the left eyebrow.
Action: she inhales, then delivers one line, eyes wet but not crying.
Context: dark studio, a vintage microphone in the lower third, out-of-focus meters behind.
Style and ambiance: warm key from the right, cool rim, shallow focus on the mouth.
Audio: she says, "I kept the chorus." Ambient: room tone, no score.
Leave headroom at the top for captions. 8 seconds.

Image: SunoMV Team · a lyric close-up is a mouth, a mic, and one quoted line

Cinematic tracking shot

Cinematography: tracking shot from the left, hip height, 50mm, modest speed.
Subject: a man in his 30s, grey overcoat, scuffed boots, a silver ring on the right hand.
Action: he walks parallel to a brick wall, left hand brushing wet ivy.
Context: alley after rain, one sodium street lamp, a cat frozen in a doorway.
Style and ambiance: orange sodium vs cold brick, puddle reflections, no beauty fill.
SFX: boot on wet stone; a distant bus hiss. Ambient: city two streets over. No score.
16:9. 8 seconds. Do not reveal the end of the alley.

Vertical short (9:16)

Cinematography: medium close-up, 9:16, slow crane-up from chest to eyes.
Subject: the same woman in the red vinyl jacket, rain in her hair.
Action: she looks up, blinks once, then smiles on the last beat.
Context: neon stairwell, one pink tube light, condensation on the rail.
Style and ambiance: magenta key, green bounce from a sign, crushed blacks.
Audio: she says, "Hit post." SFX: a single heel on metal stairs. No score.
Keep her in the center third. 8 seconds.

Character lock (same person, second location)

Cinematography: wide-to-medium, slow dolly right, 35mm.
Subject: SAME woman as previous shots — blunt black bob, oxblood leather jacket,
white tank, silver hoops, chipped black nail polish. Do not recast. Do not change hair.
Action: she leans on a convenience-store window and watches her own reflection.
Context: fluorescent store interior behind glass, night street in the reflection.
Style and ambiance: sickly green fluorescents vs magenta street neon, wet glass.
Ambient: compressor hum, distant register beep. No score. 16:9. 8 seconds.

Native-audio dialogue (Fast only)

Cinematography: two-shot, eye-level, slow pan from him to her, 40mm.
Subject A: the woman in the oxblood jacket. Subject B: a man in a grey overcoat.
Action: he turns, she does not. She answers without looking at him.
Context: rain under a bus shelter, one fluorescent tube buzzing.
Style and ambiance: hard overhead fluorescent, greenish, wet pavement sheen.
He says, "You coming?" She says, "After the chorus."
SFX: rain on the shelter roof; a bus air-brake. Ambient: traffic. No score.
16:9. 8 seconds. Pick Veo 3.1 Fast. Do not use the silent option.

Image: SunoMV Team · quoted lines only work if the mouths stay in frame

Empty prompt vs a prompt for Veo 3

Same song, same 8 seconds, two inputs. Only one of them is a prompt for Veo 3.

	Empty	Copy-paste
First clause	”cinematic, 8K”	named shot + named move
Subject	”a girl”	wardrobe + hair + one distinguishing mark
Action	”dancing”	three steps, weight, contact with the ground
Audio	(none, or “add music”)	quoted line + named SFX + “no score”
Ratio / length	”make it viral”	16:9 or 9:16, 8 seconds, start and end frames
Who it’s for	a still that pretends to be a video	a chorus cut you can actually edit

If you already have a Suno track, do not ask Veo to write another one. Point it at the picture and keep the song. The Suno prompt tips guide is the pasteable list for the audio half; this page is the pasteable list for the picture half. For a longer Seedance-style shot language, see the Seedance 2.5 prompt guide — different slots, same “copy the block” job.

Decision filter: If you cannot point to the start frame and the end frame in the prompt, you do not have a shot. You have a vibe.

FAQ

Do I write “Veo 3” or “Veo 3.1” in the prompt?

Neither. The model name does not belong in the prompt text. People search veo 3 prompt; SunoMV’s picker currently lists Veo 3.1 Fast and Veo 3.1 Lite. The five slots are the same. Put the camera in the first clause, not the version string.

Why did my duration get lifted to 8 seconds?

That is the Veo lock in SunoMV, not a bug in your wording. Write start and end frames that fit 8 seconds. Do not describe a minute of story and hope the model compresses it.

First and last frame: do I still write a prompt?

Yes. The frames hold identity. The prompt still has to name the camera, the action, and the audio. A first-last pair with cinematic 8K in the middle is still an empty prompt.

Why is there no dialogue even though I wrote a line?

Two usual causes. You picked the silent Lite option. Or the line was not in quotation marks, so it was treated as description. Quote it, pick Fast, keep the mouth in frame.

Should I write negative prompts?

Write the positive shot. Google’s samples spend their words on what is in the frame. If you must exclude something (a score, a crowd, a second face), say No score. No crowd. as a closing constraint — not a paragraph of “don’ts.”

Can one prompt be the whole music video?

No. One prompt is one 8-second shot. The AI music video creation guide is the workflow for stringing those shots. This page is only the pasteable Veo 3 prompt.

Copy, pick Veo, export the MV

A Veo 3 prompt is finished when you can paste it, not when you can explain it.

1. Copy one template above. Fill the brackets. Keep Google’s five-slot order.
2. Open the SunoMV audio-to-video generator. Drop in the song (Suno link or your own file).
3. In the model list, pick Veo 3.1 Fast if the block has dialogue or SFX. Pick Veo 3.1 Lite if you want a silent 8-second picture.
4. Paste the whole prompt into that shot. Do not split the camera sentence from the subject sentence.
5. Export the clip onto the lyric line. Repeat. The MV is the string of 8-second shots, not one heroic paragraph.

The people who get usable Veo footage are not the ones with a better adjective list. They are the ones who already have a pasteable prompt for Veo 3 before they touch generate.

SunoMV Team