Try Now

AI Video Generation

ChatCut generates video clips up to 15 seconds long using Seedance 2.0, and places them directly on your timeline. You describe a scene in natural language — including camera movement, lighting, and mood — and the AI creates it. Every generated clip includes automatic audio, so you won’t need to source sound separately.

This guide walks through the five generation modes, then shows real examples of creating commercials, brand promos, and documentary-style content.

Five Generation Modes

Depending on your starting point, you can generate video in five different ways:

  • Text to video — describe a scene from scratch and get a fully generated clip. Best for B-roll, establishing shots, and creative sequences where you don’t have source material.
  • Image to video — bring a still image to life with motion. Good for animating product photos, illustrations, or AI-generated images.
  • Character to video — animate a character with consistent identity across multiple clips. Useful for education content and explainer videos.
  • Video to video — restyle or modify existing footage. Apply new visual treatments while keeping the motion.
  • Keyframe to video — define a start frame and end frame, and the AI generates smooth motion between them. Best for controlled transitions and precise animation.

Each mode produces clips up to 15 seconds — long enough for B-roll, transitions, product shots, and social media content.

1

Choose your mode

Start from text, an image, a character reference, existing video, or keyframes

2

Describe the scene

Use natural language with camera direction: 'slow dolly forward through a neon-lit corridor'

3

Generate

Seedance 2.0 creates your clip with automatic audio generation included

4

Edit on timeline

The clip appears on your timeline. Trim, add transitions, layer with motion graphics and captions

Camera Language

You’re not limited to vague descriptions. ChatCut understands professional camera vocabulary — the same language you’d use on a film set:

  • Movement — dolly in/out, truck left/right, crane up/down, steadicam
  • Rotation — pan left/right, tilt up/down
  • Focus — rack focus, pull focus, shallow depth of field
  • Speed — slow motion, time-lapse
  • Framing — close-up, wide shot, medium shot, over-the-shoulder

Including these terms in your prompt gives you precise control over how the generated clip looks and feels. The more specific your camera direction, the better the result.

Tutorial: Cinematic Product Commercial

This example creates a high-end jewelry commercial — a good demonstration of how camera language, reference images, and detailed prompting produce cinematic results.

Write a detailed cinematic prompt

For product commercials, describe the composition, lighting, color palette, and film stock. Treat your prompt the way you’d write a shot list.

Try this prompt
A high-end jewelry commercial shot, featuring a diamond engagement ring centered on a Tiffany blue jewelry box. The ring is presented in a classic proposal composition, with the diamond facet catching the light. The background is a large window offering a captivating view of Manhattan. Natural light floods in, showcasing the cool brilliance of the diamond and the precise color of the Tiffany blue ring box. 35mm film grain, subtle distortion and halo, Tiffany blue, ivory, and cool silver tones, soft natural backlighting, moderate contrast, and slight halo effect, simulate Kodak Vision 3 format. 10 seconds, photorealistic, smooth cinematic movement.

Attach a reference image

Uploading a reference image gives the AI an anchor for composition, color, and product placement. This is especially important for product shots where accuracy matters.

Reference image of a diamond ring on a Tiffany blue box with Manhattan view
Demo
Generated Tiffany jewelry commercial — cinematic lighting and camera movement from the prompt.

Key takeaways:

  • Film stock references (Kodak Vision 3, 35mm grain) add cinematic quality
  • Color palette descriptions (Tiffany blue, ivory, cool silver) guide the AI’s color grading
  • Camera language (smooth cinematic movement) controls how the shot flows

Tutorial: App Promo with Motion Graphics

Brand promos combine AI-generated footage with motion graphics — title cards, logo animations, and branded overlays. Instead of describing every visual detail, point the agent at the brand’s website.

Let the agent analyze the brand

The agent browses the website, downloads logos and images, identifies key selling points and brand colors, then selects a motion graphics template and assembles the promo.

Try this prompt
Create a 15s high-quality promo video for spotify.com. Analyze the page to understand the product and its design style, then craft the video.
Result

The agent browsed Spotify's website, identified brand colors (green/black), downloaded the logo, and assembled a promo using the App Promo motion graphics template.

App Promo template selection in the ChatCut editor
Demo
Spotify promo — brand assets and Seedance-generated footage combined with motion graphics.

You can build the same kind of promo for any brand. Here’s an Apple promo generated with the same approach:

Try this prompt
Create a 10s promo video for apple.com. Extract visual assets (logo, original images, demo gifs/videos) from the page. Use these elements to craft the video with motion graphics. Add beat-synced background music.
Demo
Apple promo with extracted brand assets and AI-generated footage.

Key takeaways:

  • Pointing the agent at a URL lets it extract assets, colors, and messaging automatically
  • Motion graphics templates handle title cards, transitions, and branded overlays
  • AI music can be beat-synced in the same prompt

Tutorial: UGC Product Ad

User-generated content (UGC) style ads work well for e-commerce and product marketing. The agent can analyze a product listing page, understand features and selling points, write a persuasive script, and generate the ad.

Point the agent at a product URL

Try this prompt
Create a 15s UGC ad video for [product URL]. Analyze product images and key selling points from the page, write a persuasive script, and generate a scroll-stopping native ad.
Result

The agent read the product listing page, downloaded images, understood the product's key selling points, wrote a script, and assembled a UGC-style vertical ad.

Suitcase product reference from the listing page
Demo
UGC suitcase ad — scripted and edited from a product URL.

Key takeaways:

  • The agent handles research, scripting, and editing end to end
  • Product images from the URL become assets in the video
  • Portrait format (9:16) is generated for mobile-first platforms
  • Add AI voiceover narration or AI captions in the same prompt to complete the ad

Tutorial: Documentary-Style Multi-Clip Project

For longer-form content, you can generate multiple clips with different scenes and ask the agent to edit them together. This Stripe documentary promo shows the workflow.

Generate multiple scenes, then edit together

Try this prompt
Cinematic documentary-style interview, multiple professional settings: a small bakery kitchen with flour-covered counter, a dark home office with multiple monitors displaying code, a modern clean corporate office. A middle-aged woman working, a young Indian man coding, a young professional Asian man sitting at a desk. Warm, natural lighting, shallow depth of field, 35mm film aesthetic, high-end production quality, 8k resolution, authentic and emotional, shot on Alexa, soft natural light, clean background, balanced composition.
Result

We prompted a 15s documentary-style promo for Stripe, then asked for three variations with different characters. The agent generated a separate AI video for each. Once the videos were ready, we asked the agent to edit them together with music into a polished promo.

Demo
Stripe documentary-style promo — three scenes generated separately, then stitched together by the agent.

Key takeaways:

  • Generate clips individually with different settings and characters
  • Ask the agent to edit them together — it handles cuts, pacing, and transitions
  • Add background music in a follow-up prompt
  • Camera references (shot on Alexa, 35mm) and lighting descriptions (soft natural light, shallow depth of field) significantly improve cinematic quality
  • This workflow extends naturally to AI filmmaking projects

More Prompt Examples

Here are two more examples that show different styles and use cases.

Sneaker commercial

Try this prompt
A high-energy sneaker commercial for Air Jordan 6. The video alternates between fast-paced macro shots of the shoe's details (tongue, heel logo, stitching) and dynamic wide shots of a dancer's footwork. The lighting is moody and cinematic, highlighting the contrast between matte black leather and glowing red accents. Professional motion graphics feel, smooth transitions, hyper-realistic textures, 8k.
Air Jordan 6 reference photo showing shoe details
Demo
Air Jordan commercial — generated with macro and wide shot alternation.

SaaS promo

Try this prompt
Create a 10s promo video for https://lovable.dev/. Extract visual assets (logo, original images, demo gifs/videos) from the page. Use these elements to craft the video with motion graphics. Add beat-synced background music.
Demo
Lovable promo — aurora-style aesthetic matched from the website.

Working with Generated Clips

After generation, every clip lives on your timeline like any other piece of footage. You can:

  • Trim — cut the clip to the exact frames you need using the scissor tool
  • Reposition — drag clips on the timeline to reorder scenes
  • Layer — stack generated video with motion graphics, captions, and images
  • Transition — add transitions between generated clips and your own footage
  • Combine — mix AI-generated clips with uploaded footage in the same project

There’s no round-tripping between a generator and an editor. Everything happens in one workspace.

Pricing

ChatCut charges approximately 0.6 credits per second of generated video. A 10-second clip costs about 6 credits. A 15-second clip costs about 9 credits.

If a generation is rejected by safety filters, you aren’t charged. You only pay for clips you actually receive. This means iterating on prompts stays cheap — there’s no penalty for experimenting.

Combining with Other Features

Video generation is most powerful when combined with ChatCut’s other AI capabilities in a single editing session:

  • Layer motion graphics — title cards, lower thirds, and branded overlays — on top of generated footage
  • Add AI captions to make your content accessible and improve engagement
  • Generate background music that matches the mood of your video
  • Create AI voiceover narration to accompany your generated visuals
  • Use AI image generation to create thumbnails, backgrounds, or reference images for image-to-video mode
  • Apply text-based editing to refine the final cut by editing the transcript

Tell the agent everything in one prompt: “Generate a 10-second product shot, add a lower third with the product name, overlay upbeat background music, and add captions.” One instruction, multiple actions, all on the same timeline.

Ready to try it yourself?Try Now

Checking your footage...

Less editing. More creating.

It's time you had a superhuman editor on your side. ChatCut handles everything between recording and exporting.

Try it for free