Video Trimmer Online
Trim a video to the exact start and end point you need — without re-encoding, and without uploading a single byte.
How does this video trimmer work?
The trimmer runs a full build of FFmpeg compiled to WebAssembly directly inside your browser tab. Your file is written into a virtual in-memory filesystem, cut between the start and end points you chose, and written back out as a new file you can download. By default the cut is made with stream copy, which moves the existing compressed video and audio into a new container instead of decoding and re-encoding them — so it is both fast and completely lossless.
GuideHow to use it
Features & benefits
Why this runs entirely in your browser
Trimming normally means uploading a large video to someone else's server and waiting for a queue. Here, FFmpeg itself is downloaded to your browser once and the video stays put. Nothing about your footage is transmitted, so there is no upload wait, no size limit imposed by a server, and nothing to delete afterwards. The first trim of a session downloads the FFmpeg engine, and every trim after that begins immediately.
Common uses
Cutting the dead air off the beginning and end of a screen recording, isolating a single clip from a long capture, shortening a video to fit a platform's time limit, pulling one highlight out of a game or match recording, or trimming footage down before sending it to someone so the file stays small.
Frequently asked questions
A few things people usually want to know before trimming a video.
Not in the default mode. Stream copy takes the already-compressed video and audio packets from your source and writes them straight into a new container without decoding them, so every frame that survives the cut is bit-for-bit identical to the original. That is why the operation is fast even on long files — no pixels are re-compressed. Precise mode is the exception: it re-encodes the selected range with H.264, which introduces a generation of compression loss in exchange for a frame-exact starting point.
Compressed video can only begin at a keyframe, a complete self-contained frame that later frames are described relative to. When copying streams without re-encoding, FFmpeg must start the output at the nearest keyframe at or before your chosen point, which in a typical file with keyframes every two to ten seconds can shift the start earlier. If the cut has to land on an exact frame, switch on precise mode: it re-encodes the range so a new keyframe is generated exactly where you asked.
The first trim of a session downloads FFmpeg compiled to WebAssembly — roughly thirty megabytes of engine — into your browser. That download is deliberately deferred until you actually click Trim, so simply reading this page costs you nothing. Once loaded, the engine is cached in memory for the rest of your visit and by your browser's HTTP cache for later visits, so subsequent trims and the video compressor tool start working immediately.
There is no limit imposed by a server, because no server is involved, but there is a practical ceiling set by browser memory. The whole video is held in an in-memory filesystem while FFmpeg works, so files beyond roughly one to two gigabytes may exhaust the tab's available memory and fail, especially on a 32-bit browser or a low-memory device. Stream-copy trimming is far lighter than re-encoding, so if a large file struggles, keep precise mode switched off and close other heavy tabs first.
No. Your video is never uploaded, stored, or shared. It is read from your device into the browser's memory, processed by FFmpeg running as WebAssembly inside this tab, and written back out locally as a downloadable file. The only network request the tool ever makes is fetching the FFmpeg engine itself, which contains none of your data. Closing or refreshing the page discards both the source video and the trimmed result immediately.