When merging is the right move
A few obvious cases: an interview cut into Q&A chunks that needs to be one deliverable, a screen recording that auto-split into chunks because the recorder hit a duration limit, b-roll plus the main take that should ship as a single video, or multiple speakers each recorded separately on their own machine. If the source clips were created the same way, merging is fast and lossless. If they weren't, merge is the wrong tool — see below.
How it works under the hood
We use FFmpeg's concat demuxer with -c copy, which splices the bitstream of each clip end-to-end without re-encoding. That's why it's fast and lossless — but it's also why every input must share codec, profile, resolution, and frame rate. If they don't, the demuxer either errors out or produces a file with audio/video drift. We surface the error if FFmpeg complains; if you get a file back that plays weirdly, that's the diagnostic.
When you need a desktop tool instead
If your clips came from different cameras, different recording apps, or have different resolutions, stream-copy won't work. The fix is to re-encode each clip to a common format first, then merge. You can do that one-clip-at-a-time with our WebM → MP4 converter or MOV → MP4 converter and then merge — or skip the dance entirely and use HandBrake, FFmpeg locally, or a video editor on your machine. For mixed-source merges at scale, desktop is faster.
Recording in chunks on purpose?
If you're recording a long session and worried about losing it all to a crash, chunked recording is sensible — but the easier path is a recorder designed for long sessions to begin with. Clipy handles long takes natively, ships you a sharable link without an upload-step round trip, and never asks for a signup on the viewer side. Use this merge tool to tidy up after the fact; use Clipy if you'd rather not need to merge.