Video Prompts

A curated library of 10 video prompt styles covering the most common video use cases, each with copy-paste prompt variations for Google Veo 3.1 (featured), OpenAI Sora 2, Runway Gen-4.5, Kling 2.x, and Pika 2.2 — simply replace [PLACEHOLDERS] with your own content.

⚡ Quick Start with Veo 3.1

If you just want to generate something great right now:

Pick a style from the categories below.
Open the style doc and go to the Google Veo 3.1 section (listed first).
Choose the variation that matches your use case.
Copy the prompt, replace the [PLACEHOLDERS], and generate.
Iterate — adjust camera movement, lighting, pacing, and audio cues based on the result.

Have a reference image? Most platforms now support image-to-video. Veo 3.1’s “ingredients to video” accepts reference images for consistent characters and objects; Runway Gen-4 requires an input image; Kling and Pika both support image-to-video and start/end frames.

Platform Guide

Each prompt is provided in platform-specific variants. Understanding each platform’s capabilities, duration limits, and prompt conventions will help you get the best results.

🔵 Google Veo 3.1 (Featured — Best Audio + Cinematic Control)

Veo 3.1 generates video with rich, natively synchronized audio — dialogue, sound effects, and ambient noise are all driven by your prompt.

Durations: 4, 6, or 8 second clips (extendable via scene extension).
Aspect ratios: 16:9 or 9:16. Output at 720p or 1080p (4K available on some tiers).
Prompt structure: Use the five-part formula — [Cinematography] + [Subject] + [Action] + [Context] + [Style & Ambiance].
Audio cues:
- Dialogue: put specific speech in quotation marks — A woman says, "We have to leave now."
- Sound effects: describe them explicitly — SFX: thunder rumbling in the distance, rain on a tin roof.
- Ambience: set the sound bed — Ambient noise: a quiet café murmur, cups clinking.
Key features: “ingredients to video” (reference images for consistent characters/objects), “first and last frame” transitions, scene extension.
Tips: Front-load camera and shot type (“Cinematic dolly-in shot…”). Name lens and lighting like a director. Describe one continuous shot per clip for best coherence.

OpenAI Sora 2

Sora 2 generates synchronized audio and dialogue and excels at physical realism and multi-shot continuity.

Durations: 4, 8, 12, 16, or 20 seconds (set via API parameter, not prose).
Resolutions: 1280×720 / 720×1280 (sora-2); up to 1920×1080 / 1080×1920 (sora-2-pro).
Prompt structure: Write like a shot list — describe the shot type, subject, action, setting, lighting, and mood in prose. Resolution and duration are API parameters and cannot be requested in prose.
Audio cues: Include dialogue lines in quotes and describe sound design in prose; Sora 2 syncs speech and effects to the visuals.
Key features: character references (cameos) for consistent characters, video extension, remixing.
Tips: Shorter clips follow instructions more reliably. For multi-shot sequences, describe each shot in order (“Shot 1: … Shot 2: …”).

Runway Gen-4.x

Gen-4 and Gen-4.5 are image-to-video first: the input image defines composition, and the prompt describes motion.

Durations: 5 or 10 second clips at 24fps.
Aspect ratios: 16:9, 9:16, 1:1, 4:3, 3:4, 21:9.
Prompt structure: Focus almost exclusively on motion — subject action, environmental motion, camera motion, and motion style/timing. Don’t re-describe what’s in the input image.
Audio: No native audio generation — add sound in post.
Camera keywords: locked camera, handheld camera, slow dolly-in, tracking shot, orbit, crane up, crash zoom, rack focus.
Tips: Start with a minimal motion prompt and layer detail once the basic motion works. Add Continuous, seamless shot to avoid unwanted cuts. Use sequential language (“then”) for ordered actions.

Kling 2.x

Kling 2.1/2.5 produces smooth, physically plausible motion with strong prompt adherence and start/end frame control.

Durations: 5 or 10 second clips; Standard and Professional modes.
Aspect ratios: 16:9, 9:16, 1:1.
Prompt structure: Subject + Subject Movement + Scene + (Camera Language + Lighting + Atmosphere).
Audio: No native audio generation — add sound in post.
Key features: image-to-video, start/end frame control, negative prompts to suppress artifacts.
Tips: Keep motion achievable within 5 seconds — extreme camera moves in short clips look disorienting. Use negative prompts (blurry, distorted, extra limbs) sparingly and specifically.

Pika 2.2

Pika 2.2 is built around Pikaframes keyframe control — define start and end frames and let the model generate the motion between them.

Durations: 1–10 second clips at native 1080p.
Aspect ratios: 16:9, 9:16, 1:1, 4:5, 5:4, 3:2, 2:3.
Prompt structure: Short, direct motion descriptions. With Pikaframes, let the frames do the work and use the prompt to guide the transition style.
Audio: No native audio generation — add sound in post.
Key features: Pikaframes (keyframe interpolation), Pikadditions (add elements to a scene), Pikaswaps (swap objects/subjects), Pikaffects (dynamic visual effects), negative prompts.
Tips: For product or transformation shots, Pikaframes with a clean start and end image beats a long text prompt.

🎥 Camera Movement Cheat Sheet

Use precise cinematography terms — every platform responds better to named camera moves than to vague descriptions:

Term	What it does	Best for
`Static / locked shot`	Camera doesn’t move	Dialogue, product hero shots
`Dolly-in / dolly-out`	Camera physically moves toward/away	Building tension, reveals
`Tracking shot`	Camera follows the subject	Action, walking scenes
`Orbit / arc shot`	Camera circles the subject	Product showcases, hero moments
`Crane / pedestal up`	Camera rises vertically	Establishing shots, endings
`Pan / tilt`	Camera rotates horizontally/vertically	Scanning environments
`Handheld`	Natural shake and drift	UGC, documentary realism
`FPV drone shot`	Fast, fluid first-person flight	Establishing shots, adrenaline
`Rack focus`	Focus shifts between subjects	Emotional beats, reveals
`Crash zoom`	Rapid zoom toward subject	Comedy, dramatic emphasis

📐 Duration & Aspect Ratio Reference

Platform	Clip lengths	Aspect ratios	Native audio
Veo 3.1	4, 6, 8s (+ extension)	16:9, 9:16	✅ Dialogue, SFX, ambience
Sora 2	4, 8, 12, 16, 20s	16:9, 9:16 (via resolution)	✅ Dialogue, SFX, ambience
Runway Gen-4.x	5, 10s	16:9, 9:16, 1:1, 4:3, 3:4, 21:9	❌
Kling 2.x	5, 10s	16:9, 9:16, 1:1	❌
Pika 2.2	1–10s	16:9, 9:16, 1:1, 4:5, 5:4, 3:2, 2:3	❌

How to Use These Prompts

Find a style from the categories below.
Click through to the individual style doc.
Pick a variation that matches your intended output.
Copy the prompt for your platform — Veo 3.1 is listed first and recommended.
Replace placeholders — [SUBJECT], [ENVIRONMENT], [PRODUCT], etc. — with your desired content.
Iterate — adjust camera movement, lighting, pacing, or audio cues based on initial results.

🎬 Video Prompt Styles

Cinematic Sequence

Film-grade narrative shots with deliberate cinematography — named camera moves, motivated lightin...

Product Showcase

Studio-quality product reveals with orbit shots, macro detail passes, and clean controlled lighti...

Social / UGC Clip

Authentic, handheld, selfie-style content that feels like a real person filmed it on a phone — im...

Animation & Stylized

Non-photorealistic looks — 3D character animation, anime, claymation, paper cutout, pixel art in ...

Drone & Establishing Shot

Sweeping aerials, FPV fly-throughs, and slow crane reveals that set a location and scale. Establi...

Dialogue Scene

Character-driven scenes with synchronized, lip-synced speech and natural performance. Only Veo 3....

Action & Chase

High-energy pursuits, stunts, and kinetic camera work. Action is the hardest thing to prompt — ph...

Nature & Wildlife

Documentary-grade animal behavior, macro nature detail, and dramatic weather. Nature is a model s...

Timelapse & Hyperlapse

Compressed time: cities pulsing at night, clouds streaming past monuments, construction rising, s...

Brand Commercial

Polished 15–30 second spot energy compressed into single clips: lifestyle moments, brand-world mo...

Contribute! Generated a great video with one of these prompts? Share your results and prompt refinements via PR!