How AI Reframe actually works (and why it's not the same as AI Clipping)
AI Reframe is the tool people misunderstand fastest, usually because the name sits right next to "AI Clipping" in the sidebar and gets assumed to be a smaller version of it. It isn't — it's a genuinely simpler tool, and that's exactly what makes it useful. Here's what it actually does.
It reframes. It doesn't pick moments.
AI Clipping transcribes a video, reads it for the strongest 15-30 second moments, and cuts several short clips out of a long recording — see how AI Clipping actually decides what's worth posting for that pipeline. AI Reframe does none of that. It takes the entire video, start to finish, and converts it to vertical 9:16 for TikTok, Reels, and Shorts — no transcript, no moment-picking, no AI model call at all. If your source is already the right length and pacing (a talking-head reaction video, a screen recording, a clip you've already trimmed yourself), Reframe is the whole job. If it's a long recording you need cut down first, that's what AI Clipping is for.
The actual mechanics: scale up, then crop the overflow
A horizontal 16:9 recording doesn't fit a 9:16 frame no matter what you do to it, so something has to give. Reframe scales the source up until it fully covers the vertical frame, then crops the left and right overflow away, centered — the subject fills the screen edge-to-edge instead of floating in the middle with black bars on either side. That center-weighted crop is a deliberate choice: padding with blurred-background bars (common in other tools) technically preserves the whole frame but reads as "obviously repurposed" the moment someone sees it; a full-bleed crop reads as native to the platform, which matters more for whether someone keeps scrolling past it.
The honest limitation
A center-crop doesn't know what's actually important in the frame — it just cuts the same fixed amount off each side regardless of content. For a single centered speaker, that's invisible; you'd never know it was cropped at all. For two people sitting side-by-side in a wide interview shot, a center crop can genuinely cut one of them half out of frame. That's a real limitation, not a hidden one — the honest fix for that specific case is reframing the shot itself (a tighter, more centered camera angle) rather than expecting software to guess correctly which half of a wide two-shot matters more.
Why it's free on every tier
AI Clipping and AI Subtitles both call a real transcription API per video, which is the actual cost driver behind the monthly quota on both. Reframe never transcribes anything — it's pure video processing, so there's no third-party API cost to meter. That's a real cost distinction, not an arbitrary generosity: it's the same reason Reframe doesn't count against your monthly video limit at all, on any plan.
When to actually reach for it
Already-edited content that just needs a format change — a finished YouTube video you want as a vertical repost, a screen-recorded demo, a single continuous reaction clip. If you're starting from a long, unedited recording and need the software to find what's actually worth posting inside it, that's a different job — see AI Clipping vs. Reframe vs. Subtitles vs. Thumbnail for how to pick between all four in about ten seconds.