Veo 3.1 Builder

Veo 3.1 prompts with dialogue, sound and shot-by-shot control · 32 fields · 6 outputs

Library

Filled only

Subject Type

Action Beat

Emotion and Performance

Environment

Time and Weather

Speaking Setup

Dialogue Line

Voice Character

Accent and Delivery

Ambient Bed

Sound Effects

Shot Size

Camera Movement

Lens

Aperture and Depth

Camera Body Look

Frame Rate and Motion

Lighting Setup

Color Palette

Film Stock and Grade

Texture and Artifacts

Atmosphere

Veo Model

Current Veo model (with Veo 3.1 Fast and Veo 3.1 Lite). Native clips are 4, 6 or 8 seconds; longer videos are built with scene extension.

Generation Mode

Duration

Aspect Ratio

Resolution

Reference Images

Consistency Lock

Cinematography Style

Negative Prompt

Delivery Platform

🔒 Preview mode
Try every field. Unlock Veo 3.1 with Pro to copy and save the prompts.

✨ Enhance with AI
Rewrite the main prompt into 3 polished variations. Pro feature.

01Natural Language Prompt

Medium close up of a solo character speaking directly to camera with calm restraint, in a rain slicked city street at golden hour. Slow push in on a 50mm standard lens at T2.0 Shallow, shot on ARRI Alexa 35 at 24fps cinematic. Lit with single source window, teal and orange palette, graded as ARRI LogC Natural, fine 35mm grain and light haze. Naturalistic available light style. Single speaker to camera, in a warm male mid voice with a Neutral American accent. Audio: room tone only, footsteps on concrete. Avoid warped faces.

02Main Veo Prompt

SUBJECT: Solo Character - Speaking Directly to Camera - Calm Restraint SCENE: Rain Slicked City Street during Golden Hour DIALOGUE: Single Speaker to Camera VOICE: Warm Male Mid with Neutral American delivery AUDIO: Room Tone Only bed with Footsteps on Concrete FRAMING: Medium Close Up CAMERA: Slow Push In LENS: 50mm Standard at T2.0 Shallow shot on ARRI Alexa 35 at 24fps Cinematic LIGHTING: Single Source Window COLOR: Teal and Orange graded as ARRI LogC Natural TEXTURE: Fine 35mm Grain with Light Haze STYLE: Naturalistic Available Light NEGATIVE: Warped Faces

03Native Audio and Dialogue Script

NATIVE AUDIO SCRIPT - VEO GENERATES ALL SOUND SETUP: Single Speaker to Camera VOICE PROFILE: Warm Male Mid ACCENT AND READ: Neutral American PERFORMANCE: deliver the line with Calm Restraint LIP SYNC: keep mouth movement locked to the spoken line with no drift AMBIENT BED: Room Tone Only present under the dialogue at low level SOUND EFFECTS: Footsteps on Concrete triggered by the action ROOM ACOUSTIC: match reverb to a Rain Slicked City Street space NO MUSIC unless requested - Veo scores the scene from the description AVOID: Warped Faces

04Camera and Lens Brief

SHOT SIZE: Medium Close Up MOVEMENT: Slow Push In LENS: 50mm Standard STOP: T2.0 Shallow BODY: ARRI Alexa 35 FRAME RATE: 24fps Cinematic BLOCKING: the subject performs Speaking Directly to Camera inside the frame FOCUS: hold critical focus on the eyes throughout the move LIGHT DIRECTION: Single Source Window ATMOSPHERE: Light Haze catching the light ARTIFACTS: Fine 35mm Grain ONE MOVE ONLY - do not combine a second camera move in this clip

05Reference and Consistency Brief

MODE: Text to Video REFERENCE INPUT: Single Character Reference LOCK: Face and Expression Lock must hold across every generation IDENTITY: keep the Solo Character facial structure and wardrobe identical to the reference LOCATION: keep Rain Slicked City Street layout and background objects in the same positions GRADE LOCK: repeat Teal and Orange and ARRI LogC Natural verbatim in every shot LIGHT LOCK: keep Single Source Window coming from the same direction FORMAT: 16:9 Landscape at 1080p Full HD native - do not crop from another ratio MODEL: Veo 3.1 for 8 Seconds DRIFT CHECK: reject any take where the face or wardrobe shifts mid clip

06Five Shot Sequence

FIVE SHOT SEQUENCE - EXTEND CHAIN STYLE LOCK: Naturalistic Available Light with Teal and Orange and ARRI LogC Natural on every shot CONSISTENCY: Face and Expression Lock carried through the whole chain SHOT 1 ESTABLISH: wide of Rain Slicked City Street during Golden Hour no dialogue SHOT 2 INTRODUCE: Medium Close Up on the Solo Character beginning to Speaking Directly to Camera SHOT 3 SPEAK: close up delivering "" in a Warm Male Mid voice SHOT 4 REACT: cut to the reaction carrying Calm Restraint SHOT 5 RESOLVE: Slow Push In away leaving the Rain Slicked City Street in frame LENS THROUGHOUT: 50mm Standard at T2.0 Shallow on ARRI Alexa 35 AUDIO CONTINUITY: Room Tone Only runs unbroken across all five shots EACH SHOT: 8 Seconds at 16:9 Landscape NEGATIVE: Warped Faces DELIVER: cut for YouTube Shorts

📘 Pro Guide

Veo Model Guide

Veo 3.1Current flagship. Best prompt adherence and audio qualityDefault choice
Veo 3.1 FastSame family, faster turnaround, slight quality tradeIteration passes
Veo 3.1 LiteLowest cost tier for high volume workBulk social content

Native Audio Rules

Veo generates all soundDialogue, effects and ambience are produced with the video, not added afterCore difference
Quote the exact linePut the spoken words in quotation marks or Veo will paraphrase themCritical
Keep lines shortUnder about twelve words per clip keeps lip sync tightLip sync
Describe the voiceAge, timbre and accent all steer the generated voiceCasting
Name the roomVeo matches reverb to the described spaceAcoustics
Two speakers is the limitThree or more voices in one clip degrades badlyScene design

Lens and Depth Reference

18mm to 24mmWide environmental framing, visible edge distortionEstablishing
35mmNatural reportage feel, good for two handersConversation
50mmClosest to human vision, neutral and versatileDefault
85mm to 135mmCompressed flattering portraits, strong background meltClose ups
200mmHeavy compression, isolates subject from backgroundDramatic
AnamorphicOval bokeh, horizontal flares, wide scope framingCinematic look
T1.3 to T2.0Very shallow, only one plane sharpIsolation
T5.6 to T11Deep focus, foreground and background both readableScale

Prompt Writing Rules

One camera move per clipTwo moves in a single generation produces mushSplit the shot
Lead with the subjectSubject and action before any technical directionOrder matters
Repeat the grade verbatimCopy the same stock and palette string into every shotConsistency
Always set a negativeWarped faces and lip sync drift are the usual failuresQuality control
Shoot vertical nativelyVeo 3.1 generates 9:16 directly, never crop from 16:9Social
Reference images beat textA character reference holds identity far better than descriptionCharacter lock

Model status (checked Nov 2026)

Veo 3.1Current Veo model (with Veo 3.1 Fast and Veo 3.1 Lite). Native clips are 4, 6 or 8 seconds; longer videos are built with scene extension.Veo 3.0 and Veo 2 were retired from the Gemini API on 30 June 2026.
Gemini Omni FlashSince Google I/O (May 2026) the Gemini app generates video with Gemini Omni Flash instead of Veo. Veo 3.1 stays available in Flow, the Gemini API and Vertex AI.