PH Prompt Helper

AI Video Prompt Builder

Build AI video prompts that fit the model you pick.

Pick a model, describe the video, and copy a Text or JSON prompt sized to that model's clip lengths and aspect ratios. Free, in your browser, no signup.

Open the builder Choose a model
Live prompt console Open
UGC Ad 9:16 / dialogue Product Orbit 16:9 / studio Dialogue Scene 16:9 / dialogue Cinematic Reveal 16:9 / cinematic
UGC 8s vertical testimonial 9:16 / 8s / no captions / natural audio
Generated Text 
Selfie video of a creator holding a skincare bottle in a bright bedroom. Casual hook, natural handheld motion, no captions.
Customize Copy-ready

Models

Start with the model you actually run.

Open the builder

Every video model has its own clip lengths, aspect ratios and prompting habits. Open the builder for yours: it is preset to that model's limits, with a short guide on how to prompt it.

Model limits checked . Models change often; each builder page lists its sources.

Prompt formats

Start with the video you want, not a blank prompt box.

Build one now

Each format preloads a shot structure (subject, camera, light, sound) that you can edit and send to any model above. They are starting points, not guaranteed outputs.

Animating a still image? Open Veo image-to-video mode.

9:16 / dialogue UGC Ad

Selfie video of a creator holding a skincare bottle in a bright bedroom. Casual hook, natural handheld motion, no captions.

 16:9 / studio Product Orbit

A matte-black earbuds case on a seamless studio surface. Slow orbit, crisp edge highlights, final centered hero frame.

 16:9 / dialogue Dialogue Scene

Two coworkers at a cafe table, one leans in and delivers a short line. Natural timing, room tone, no subtitles.

 16:9 / cinematic Cinematic Reveal

A lone astronaut crosses a red desert toward a crashed spacecraft. Wide shot, drone flyover, sparse orchestral score.

Prompt library

Copy a prompt, then tailor it in the builder.

View library

Every prompt carries a verification label: owner-tested, community-verified with a linked source video, or untested. The library covers Veo and Gemini Omni today; Kling, Seedance and Grok Imagine are covered by their builder guides.

Gemini Omni Flash Gemini Omni Prompts — Copy-Paste Prompts for Gemini Omni Flash

Copy-paste prompts for Gemini Omni Flash, Google's multimodal video model. Image, audio and video inputs, conversational editing, and native sound. 14 examples.

 Veo Veo Prompt Theory Prompts — AI Characters Who Know They Are AI (Copy-Paste)

Copy-paste Veo prompts for the Prompt Theory trend: AI-generated characters who realise they are AI. The exact structure behind the viral format, 12 examples.

 Veo Veo Dialogue & Audio Prompts — Make Characters Speak (Copy-Paste)

Veo dialogue and audio prompts: lip-sync, SFX, and ambient sound. Learn the says-quote syntax and copy 20 examples, 10 with public output evidence.

 Veo Veo Image-to-Video Prompts — Animate Any Image (Copy-Paste)

Veo image-to-video prompts. The key: describe only the motion, camera, and audio, not the scene the image shows. 18 examples, 8 with public evidence.

 Veo Veo UGC-Style Prompts — Selfie Vlog, Testimonial & Street (Copy-Paste)

Veo UGC prompts: selfie vlogs, testimonials, street interviews, and ASMR with a handheld look and native audio. 18 examples, 8 with public evidence.

 Veo Veo Prompts for Product Video — Copy-Paste Examples

Veo product video prompts for ads: hero shots, 360° spins, and lifestyle scenes. 17 copy-ready examples, 10 with public output evidence.

Common questions

What Prompt Helper does, in plain terms.

Read the guides
What is an AI prompt helper?
An AI prompt helper turns a plain description of what you want into a structured prompt a model can follow. Prompt Helper does this for AI video: you describe the subject, setting, camera and audio, pick a video model, and it assembles a prompt with the parts that model responds to, sized to its clip lengths and aspect ratios, instead of leaving you to guess at phrasing.
Is Prompt Helper free to use?
Yes. The prompt builder runs entirely in your browser and is free with no account. Only the optional AI Brief Assist, which drafts the brief fields for you, requires signing in.
Which AI video models does it support?
Five today: Veo (3.1, 3.1 Fast and 3.1 Lite), Gemini Omni (Omni 1.1 Flash), Kling (3.0 and 3.0 Omni), Seedance (2.5, 2.0 and 2.0 Fast) and Grok Imagine (Video 1.5). Each has its own builder page with that model's limits and prompting notes, and more models are added in small batches after each one is checked against the vendor's own documentation.
Does the same prompt work on every model?
Not reliably. Models differ in clip length (from 4 to 8 seconds on Veo, up to 15 on Kling and Grok Imagine, up to 30 on Seedance 2.5), in supported aspect ratios, and in how they treat negative prompts: Veo, Gemini Omni, Kling and Seedance take them inside the prompt, while for Grok Imagine the builder leaves the negative line out and suggests phrasing what you want positively. Choose the model first and the builder adjusts the prompt to it.
What is the difference between the Text and JSON output?
Text is the paste-ready prompt for a video app or web interface. JSON is the same prompt broken into named fields (subject, action, camera, lighting, audio, negatives), which is easier to version, diff and reuse across a batch of shots or feed to an API.
Do I need prompt engineering experience?
No. Every format starts from a working structure you can copy and adapt, and every prompt in the library shows whether it has been tested. If you want the reasoning behind the structure, the guides cover the prompt formula, camera language, and how negative prompts fix common failures.