Skip to main content

AI Video Prompts

Find your next idea. Explore prompts for video, images and sound, then copy a starting point and make it your own.

Prompt Library

Browse AI video prompts

Explore video, image and audio ideas by tool. Preview our original examples and copy a prompt to make it your own.

174 ideas to explore
174 prompts
EditingSuperVids example
Add Audio to Video

An evening walk, with a soundtrack

SuperVids prompt

Reuse the original SuperVids rainy-evening film. Add the original acoustic music at 13% volume under the existing voice, trim music to the video length and fade it out over the final 0.7 seconds. Copy the video stream without re-encoding; mix the two audio tracks with FFmpeg. No new video generation. Original source: /media/prompt-library/generated/home-rainy-evening-h3.mp4

EditingSuperVids example
Remove Audio from Video

The rooftop dance, without sound

SuperVids prompt

Reuse the original SuperVids rooftop dance. Copy the video stream into MP4 and remove every audio stream with FFmpeg -map 0:v:0 -c:v copy -an. Preserve the frame sequence, resolution and motion. This is actual audio removal, not a muted player. Original source: /media/prompt-library/generated/home-rooftop-dance-h3.mp4

EditingSuperVids example
Remove Bad Takes

Keep the coffee creator’s final take

SuperVids prompt

Reuse the original SuperVids coffee creator recording. Remove the opening false start “So, um, let me start again.” Retain source seconds 3.8–8.0: “One little coffee. My favorite way to start the day.” Cut picture and sound together using FFmpeg; preserve the pause between the retained sentences. This is a manually reviewed illustrative edit, not an automatic tool export. Original source: /media/studios/generated/coffee.mp4

EditingSuperVids example
Remove Silence from Video

The same words, shorter pauses

SuperVids prompt

Reuse the original SuperVids coffee recording. Keep source intervals 0.20–1.82, 2.15–3.40, 3.85–5.40 and 5.55–7.65 seconds with synchronized picture and audio cuts. Retain every spoken word, including the false start. Shorten quiet gaps and leading/trailing room tone. Use FFmpeg trim, atrim and concat; do not change speech speed. Manually reviewed illustrative edit. Original source: /media/studios/generated/coffee.mp4

EditingSuperVids example
AI Voiceover for Video

A voice for the morning ritual

SuperVids prompt

Reuse the original espresso product insert. Generate Brian narration with ElevenLabs Multilingual v2 directly: “One little coffee. A warmer start to your day.” Stability 0.5, similarity 0.75, speed 1. Start narration 0.9 seconds into the shot. Mix it over the original quiet machine sound using FFmpeg. Do not alter the product picture. Produced as a showcase asset, not through the static voiceover form. Original source: /media/studios/generated/coffee-product.mp4

EditingSuperVids example
Lip Sync AI

A new line, synchronized lips

SuperVids prompt

Reuse the original SuperVids fictional presenter. Generate Brian speech with ElevenLabs Multilingual v2 directly: “Start with a clear idea. Choose your scene, then bring your story to life.” Submit the original video and this new voice track to volcengine/video-to-video-lip-sync, lite mode, vocal separation off, looping off. The output follows the target audio length. This is actual lip-sync inference, not an audio-only replacement. Original source: /media/subtitles/generated/lesson-original.mp4

EditingSuperVids example
AI Video Enhancer

A closer look at the perfume shot

SuperVids prompt

Submit the original SuperVids perfume-reveal video to Kie video/upscale with resolution “2K”. Use the unmodified 1344 × 768 source: do not manufacture a blurry before image. The returned video is 2520 × 1440 and 5.17 seconds. Preserve the product and camera sequence. Review the returned video at full resolution for edge clarity and artifacts. This is provider-processed showcase footage, not an export from the static enhancer form. Original source: /media/prompt-library/generated/home-perfume-reveal-h3.mp4

EditingSuperVids example
Video Compressor

The forest film, easier to share

SuperVids prompt

Reuse the original SuperVids forest-stream MP4. Encode H.264 with libx264 CRF 29, slow preset, AAC 96 kbps and MP4 faststart. Preserve frame dimensions and duration. Source 3,123,272 bytes; result 1,483,894 bytes. Actual reduction 52.5%; review fine foliage and stream motion for compression tradeoffs. Produced offline with FFmpeg. Original source: /media/prompt-library/generated/image-video-forest-h3.mp4

Editing

A different angle can turn an ordinary moment into a story.

SuperVids example
Video to Audio

Take the evening-walk soundtrack with you

SuperVids prompt

Extract the audio from the original SuperVids rainy-evening video with FFmpeg -vn -c:a copy into M4A. Preserve the existing voice and ambience; do not synthesize new speech or separate stems. Also encode a 192 kbps MP3 listening copy. The library audio is the lossless stream-copy M4A export. Original source: /media/prompt-library/generated/home-rainy-evening-h3.mp4

CinematicSuperVids example
AI Drama Generator

The woman in blue

SuperVids prompt

Use the exact supplied woman with short dark bob and cobalt blue coat. Eight second cinematic scene, one continuous take. She walks slowly forward along the rainy street toward the amber cafe window. Camera tracks behind her left shoulder, pushes closer to her three-quarter profile as she slows and breathes, ending on a quiet contemplative close-up against warm bokeh. Preserve face, coat, street and weather. Natural rain and distant city sound, no speech, music, text or cuts.

FantasySuperVids example
AI Drama Generator

Beyond the gates

SuperVids prompt

Eight-second cinematic continuous shot. Follow the two riders on horseback slowly forward through the open bronze gates of this exact limestone castle. Low camera tracks behind the riders then smoothly cranes upward to reveal courtyard and slate towers. Preserve architecture, gate geometry and each horse and rider. Natural hoofsteps, soft wind and distant birds. No dialogue, music, captions, sudden cuts or morphing.

CreatorSuperVids example
Editing Studio

Morning ritual: the original take

SuperVids prompt

One continuous eight-second authentic coffee creator recording in this exact kitchen. Preserve her face, cream shirt, cup and espresso machine. Locked camera, natural morning light. She looks at camera and speaks with a warm natural American female voice: at about one second "So, um, let me start again." Small embarrassed smile and pause. At about three seconds "One little coffee." Pause while raising the cup slightly. At about five seconds "My favorite way to start the day." Smile to finish. Accurate synchronized lips, natural hands, no extra words, no subtitles, no music, no cuts. Quiet room tone.

Showing 12 of 174 prompts

How to write a good AI video prompt

  • Subject: who or what is in the shot.
  • Action: one clear thing that happens.
  • Camera: shot size and movement.
  • Style and light: mood, era, color.
  • Sound: ambience or a line of dialogue, if the model supports audio.

AI video prompts by model

Each model reads prompts differently. See Kling 2.6, then try your prompt in the AI video generator or start from text with text to video.

AI video prompts FAQ

Can I use these prompts for free?

Yes. Copy any prompt and use it anywhere.

Why does the same prompt look different?

Models and random seeds vary. Generate a few takes and keep the best.

Try a prompt now

Open the generator