How to change video aspect ratio with AI: step-by-step
Quick answer
To change a video’s aspect ratio with AI, upload it to Magic Hour’s AI Video Expander, open the Output menu that appears after the video loads, choose a different ratio and click Expand Video. Preview the generated edges before downloading. The guest tool offers three free previews per day, up to five seconds each. Use expansion when you need new surrounding image area; use a crop or borders when you can keep the existing pixels.
Magic Hour publishes this guide and includes its own product in the comparison. Treat the recommendations as editorial guidance from a vendor, and verify the linked first-party product details and your own output requirements before choosing a tool.
Crop, resize, reframe or expand: what changes?
Aspect ratio is width divided by height. Changing 1920 × 1080 to 1280 × 720 keeps the same 16:9 shape. Changing it to 1080 × 1920 creates a 9:16 canvas, so you must decide what fills the extra space or what gets cropped.
Method | What changes | When it fits |
|---|---|---|
Crop or auto reframe | Removes outer areas; auto reframe moves the crop to follow a subject | A speaker or product fits inside the smaller frame |
Fit with borders | Scales the complete frame into a new canvas; fills unused space with a color or background | Every original detail must remain visible without invented scenery |
AI expansion / outpainting | Generates new visual content outside the original frame | You want a filled canvas and can review the generated surroundings |
Resize / scale | Changes pixel dimensions | You need a different resolution at the same shape |
AI extension | Generates additional time after a clip | The video needs more seconds, not a different shape |
Expansion aims to keep the original composition while filling the new area. It cannot recover the true scene that was outside the camera’s view. Inspect the result for altered details as well as errors in the added area.
Expand a short video into a new aspect ratio
Upload a representative clip, choose the target shape, and inspect both the original frame and the generated edges before exporting.
Try AI Video ExpanderWhich Aspect Ratio Should You Choose?
Choose the ratio for the specific placement, not just the platform name. These are common starting points, not a complete list of allowed upload formats.
Destination | Common starting ratio | Example canvas |
|---|---|---|
Full-screen vertical short, Reel or Story | 9:16 | 1080 × 1920 |
Widescreen video or presentation | 16:9 | 1920 × 1080 |
Square feed creative | 1:1 | 1080 × 1080 |
Portrait feed creative | 4:5 | 1080 × 1350 |
A 16:9 landscape clip can become 9:16 in several ways. If one speaker fits inside a vertical crop, auto reframing may be enough. If both edges contain important action, keep the full frame with borders or generate new space above and below it. Leave room for the destination’s captions and interface overlays; a ratio alone does not guarantee they will stay clear.
What You Need
Before you begin, make sure you have:
- A video file you have permission to edit.
- A modern browser.
- The output shape required by your placement.
- A short representative section to evaluate before processing a longer clip. The guest preview uses up to the first five seconds; trim the file beforehand if you need to assess a later section.
Tools for different aspect-ratio jobs
Tool | Relevant workflow | What to inspect |
|---|---|---|
Generates surrounding image area with video outpainting | Added objects, borders, motion consistency and source details | |
Detects subjects and adjusts framing | Whether the chosen crop retains every important speaker or object | |
Changes clip ratios with editor, caption and template options | Subject framing and caption placement in each output | |
Preset or custom canvas sizes, scaling, positioning and trimming | Crop boundaries and unused canvas space |
These are different editing operations, not interchangeable quality rankings. Magic Hour is the relevant option here when you want generated surroundings. The linked product descriptions and Magic Hour’s browser controls were checked on September 10, 2026.
How to Change a Video's Aspect Ratio With AI: Step by Step
Step 1: Open Magic Hour's AI Video Expander
Open Magic Hour’s AI Video Expander. The guest form starts with an upload control. Its Presets menu contains sample clips; it is not the aspect-ratio menu.
Step 2: Upload Your Video
Click Upload video and choose your clip. Wait for it to load. The guest workflow previews up to the first five seconds; use the full video expander for longer clips and additional controls.
Step 3: Select Your Target Aspect Ratio
Once the source loads, open Output and choose the target ratio. The guest menu includes 9:16, 16:9, 1:1, 4:5, 4:3, 3:4, 3:2 and 21:9. Choose a ratio different from the source. In the full editor, the control is labeled Aspect Ratio and also includes Custom.
Step 4: Generate the Expanded Video
Click Expand Video once, then wait for the current request to finish. Processing time varies; the result is a wider or taller frame, not a longer video.
Step 5: Preview and Download
Watch the whole finished preview. Check faces, product labels and moving objects in the original area, then inspect the join and generated surroundings. Download the result if it is usable, and review it in the destination’s upload preview for cropping or covered captions.
Free video-expander results may include a watermark. Paid plans remove video watermarks and permit commercial use; check the current plan details before preparing client work.
Tips for Better AI Video Expansion
- Test a representative shot. Include the hardest head turn or camera movement, not only a still opening.
- Work around scene cuts. Separate very different shots if one expansion produces inconsistent surroundings.
- Inspect source details. Product labels, hands and faces need checking even when the goal is to preserve the original image.
- Use borders for exact visual evidence. A screen recording or documentary shot may need a clearly separate background rather than invented surroundings.
- Check the final placement. Captions, usernames and playback controls can cover content even at the correct ratio.
What to Try Next
If the clip also needs more time, see how to extend a video with AI. To transcribe speech into captions, use the Subtitle Generator, then check text placement for your output shape. For YouTube delivery settings, see the video format guide.
Frequently Asked Questions
Yes. The public tool allows three free previews per day, up to five seconds each, without an account. Free video results may include a watermark. Longer clips and additional controls are available through the full editor; paid plans include watermark-free video exports and commercial use.
Outpainting is intended to keep the original frame visible and generate surrounding content. That is different from cropping, but it is not a guarantee that every detail will remain unchanged. Compare the result with the source, especially faces, text and products.
The guest Output menu has 9:16, 16:9, 1:1, 4:5, 4:3, 3:4, 3:2 and 21:9. It appears after a video has loaded. The full editor has these presets plus a Custom option.
Commercial use is available with Magic Hour's paid plans. The free version is intended for testing and personal experimentation.
AI Video Expander changes the shape of your video by generating new content around the edges of each frame. AI Video Extender solves a different problem by generating additional footage at the end of your clip, making the video longer rather than wider or taller.
Why does it ask for a different aspect ratio?
The expander needs a different output shape to add surrounding image area. If your source is already 9:16 and you select 9:16, use a standard resize or upscale workflow when your actual goal is more pixels. Changing resolution and changing aspect ratio are separate operations.
Runbo Li is the Co-founder and CEO of Magic Hour, where he builds AI video and image tools for content creation. He is a Y Combinator W24 founder and former Data Scientist at Meta, where he worked on 0-1 consumer social products in New Product Experimentation. He writes about AI video generation, AI image creation, creative workflows, and creator tools.
