By the Recapo.ai Editorial Team · Fact-checked July 10, 2026
Auto-reframe is how you turn a horizontal 16:9 recording into a vertical 9:16 clip without hand-drawing the crop on every scene — and the version worth using isn't a static center crop, it's AI subject tracking that follows the person on screen so they stay in frame the whole way through. This guide shows you how to auto-reframe a video to vertical right in a browser, what separates a good smart reframe from a lazy one, and where the automation still needs a human eye. If you post to TikTok, YouTube Shorts, or Instagram Reels, reframing isn't a one-off chore — it's a step you repeat on every clip you ship, so it pays to get the workflow right once.
Here's the core distinction most "convert horizontal video to vertical" tutorials skip: changing the aspect ratio is the easy part. Keeping the right thing inside the new, narrower frame is the hard part — and it's the whole reason "auto reframe horizontal to vertical" exists as a feature instead of a one-click resize. Once that clicks, every tool choice below gets simpler.
What auto-reframe actually does
A 16:9 frame is wide; a 9:16 frame is tall and narrow. When you go from one to the other, roughly two-thirds of the horizontal picture has to be cut away. The only question that matters is which two-thirds.
A static crop picks a fixed window — usually dead center — and holds it for the entire clip. If your subject never moves and always sits in the middle, that's genuinely all you need. But the moment your speaker walks, leans, gestures to the side, or a second person enters, a static crop leaves them half-cut or drifting toward the edge.
An AI reframe to 9:16 works differently. It detects the salient subject in each moment — a face, a body, the main area of motion — and moves the crop window over time to keep that subject centered. Some call it smart reframe or subject tracking; the mechanism is the same idea: the frame follows the action instead of forcing the action to sit still. For talking heads, interviews, cooking, gameplay, and anything where the subject moves, that tracking is the difference between a clip that looks natively vertical and one that looks like a lazy crop of a widescreen file.
Static crop vs. AI subject-tracking reframe
Neither approach is "better" in the abstract — they fit different footage. Here's how to tell which one your clip needs.
| Static center crop | AI subject-tracking reframe | |
|---|---|---|
| How it picks the frame | Fixed window, usually center | Follows the detected subject over time |
| Best for | Subject stays centered and still | Speakers who move, gesture, or shift |
| Two people / cutaways | Often cuts someone off | Can track and re-center on the active subject |
| Effort | One setting | Auto first pass, then quick review |
| Main risk | Subject drifts out of frame | Tracking locks onto the wrong subject |
If your footage is a locked-off shot of one person in the middle, don't overthink it — a plain vertical crop is all you need, and our guide on how to make a video vertical walks through that simpler path. Come back to auto-reframe when the subject moves, because that's exactly the case a static crop handles badly.
How to auto-reframe a video to vertical (step by step)
The whole thing runs in a browser tab — no install, and no desktop app to keep updated. Here's the loop end to end.
- Upload your horizontal file. Drop in your 16:9 MP4 or MOV — a talking-head recording, interview, stream VOD, or webinar. Browser-based editors like Recapo accept common formats up to 6GB total per task, which covers most single recordings.
- Set the output to 9:16. Choose vertical as your target aspect ratio so the canvas becomes 1080×1920, the standard for TikTok, Shorts, and Reels.
- Turn on subject tracking. Let the AI make its first pass at the crop path, following the speaker or main action across the clip instead of holding a fixed center window.
- Scrub through and fix the drift. This is the step people skip and regret. Watch the reframed clip once at speed, and wherever the frame locks onto the wrong subject or lags a fast move, nudge the crop position. Ninety seconds of review saves a re-export.
- Add captions. Some vertical feeds autoplay on mute, so burned-in captions improve accessibility and comprehension. Generate them from the transcript and check the wording against the audio.
- Design a cover and export. Set a thumbnail frame, then export a clean 9:16 file ready to post.
That's the entire job, and once you've done it once, it's the same five-minute loop for every future clip.
What a good vertical reframe keeps in frame
Auto-reframe gets you most of the way; the last stretch is knowing what the frame should hold. Use this quick checklist when you review the AI's pass:
- Eyes in the upper third. Faces read best when the eyeline sits high in the tall frame, not dead center.
- Keep the action, not just the face. For cooking, unboxing, or tutorials, the hands and the object matter as much as the head — make sure tracking didn't crop them out.
- Protect on-screen text. If your source has lower-thirds, scores, or captions baked in, a tight vertical crop can slice them. Reposition or re-add them.
- Re-center on scene changes. After a hard cut, the important subject may be on the other side of the frame. Confirm the reframe jumps rather than drifts.
- Leave a little headroom. A sliver of space above the head looks intentional; a scalp pinned to the top edge looks cropped.
None of this is exotic — it's just the eye a static crop can't bring, which is why the human review step stays in the loop even with good automation.
Where auto-reframe still needs you
Being honest about the limits saves you from shipping bad clips. Subject tracking is strong on single, clear subjects and gets shakier as the scene gets busier:
- Two or more people talking. The tracker has to guess who's active; on fast back-and-forth interviews it can pick wrong. Split the clip by speaker, or key the framing manually on those beats.
- Wide group or crowd shots. There's no single subject to follow, so the crop has nothing clean to lock onto. Sometimes a static wide-to-vertical crop actually looks better here.
- Rapid cross-cutting or heavy motion. Sports and fast montages can outrun the tracker, producing a jittery frame. Slow the pacing or correct the worst cuts by hand.
- Graphics-heavy footage. Slides, charts, and gameplay HUDs weren't shot for vertical; you'll often need to reposition or rebuild them rather than trust the crop.
The rule of thumb: the clearer your single subject, the more you can trust the automation; the busier the frame, the more you should review it. Auto-reframe is a first draft that's right most of the time, not a fire-and-forget button.
Reframe as a repeatable step in your shorts workflow
Here's the part that actually matters for anyone doing this at volume: reframing is never the whole job. It's one link in a chain you run on every upload — clip the long recording, reframe each clip to vertical, caption it, export it. Doing those in four disconnected apps is where the hours leak.
In one browser workspace the chain stays together:
- Clip first. Use an AI tool to turn long videos into short clips to surface the strongest 20–60 second moments from a full recording — our guide on making YouTube Shorts from long videos covers how to spot the ones that travel.
- Reframe each clip to vertical. Auto-reframe the selected moments to 9:16 so they fill the screen instead of sitting letterboxed.
- Match the target platform. Resize for the exact feed you're posting to — a TikTok video resizer or an Instagram Reels resizer confirms the dimensions land right before you export.
- Caption and export. Burn in captions and export each clip.
Recapo bundles this exact loop — clip, reframe to vertical, transcribe and caption, add AI voiceover, design a cover, and export — in a browser, with nothing to install and up to 6GB total per task. It's not a color-grading suite or a frame-level timeline, and it won't pretend to be; what it removes is the app-switching tax on the repurposing loop you run every week. Pricing lives on the pricing page.
How to pick an auto-reframe tool
Every converter page in this search claims a clean vertical crop. Instead of trusting the demo reel, run one controlled test with your own footage — it's the only thing that tells you the truth for your workflow.
- Pick one representative file. Not a clean promo clip — a real recording where your subject actually moves, ideally with a second person or some gesturing.
- Auto-reframe it to 9:16 in each tool you're weighing, using the same source.
- Judge the tracking, not the marketing. Watch the exported vertical clip and ask: does the subject stay centered when they move? Does it re-center after cuts? Does it ever lock onto the wrong thing?
- Check the export directly. Confirm resolution, aspect ratio, and whether the free tier adds a watermark by inspecting the actual file — don't take a landing page's word for it.
- Kill anything that drifts. If the subject wanders out of frame and there's no easy way to correct it, that tool will cost you time on every clip. Move on.
This is deliberately framework-not-verdict: prices, quotas, and export limits change, and you'd have to confirm them yourself anyway. The test above answers the one question that matters — does its tracking hold your subject in frame — for the footage you actually publish.
FAQ
What does "auto-reframe" mean?
Auto-reframe converts a video from one aspect ratio to another — most often horizontal 16:9 to vertical 9:16 — while automatically repositioning the frame to keep the important subject in view. The "auto" part is AI subject tracking: instead of a fixed center crop, the frame follows the speaker or main action across the clip so they don't drift out of the narrower vertical frame.
Is AI reframe better than a manual crop?
It depends on your footage. For a still, centered subject, a manual vertical crop is simpler and just as good. AI subject-tracking reframe earns its keep when the subject moves, gestures, or shares the frame with someone else — cases where a fixed crop would cut them off. The honest test is to run both on a clip where your subject actually moves and compare the exports.
What resolution and aspect ratio should a vertical clip be?
The standard for TikTok, YouTube Shorts, and Instagram Reels is 9:16 at 1080×1920. Some feeds also accept 4:5 for in-feed posts, but 9:16 is the full-screen vertical target. Always confirm current specs on each platform's official creator page, since supported sizes change.
Can I auto-reframe a video to vertical in a browser?
Yes. Browser-based editors run the whole reframe in a tab with nothing to install — upload your horizontal file, set 9:16, let subject tracking make its pass, review the drift, caption, and export. That keeps reframing on the same surface as clipping and captioning, which is where the time savings come from if you do this often.
Why does my reframed clip cut off the subject?
Usually because a static crop is holding a fixed window while your subject moves, or because subject tracking locked onto the wrong element in a busy frame. Switch to a subject-tracking reframe if you were using a fixed crop, and on multi-person or graphics-heavy clips, review the crop path and correct the beats where it picks the wrong subject.
Reframe your first clip
The fastest way to see whether AI reframe fits your footage is to run one real clip through it. Take a horizontal recording where your subject actually moves, create a free Recapo account, reframe it to 9:16, caption it, and export — then judge the tracking against a plain center crop of the same file. If your work is vertical shorts at volume, you'll feel the difference on the first clip, because it's the exact step you'll repeat on every one after it.
Top comments (0)