Video Trimmer + Converter — Free, Online, No Upload

Apps / Video Trimmer & Converter

Drag & drop a video here, or pick a file. Supported: MP4, MOV, WebM, MKV, AVI, M4V, FLV. Up to a few hundred MB works comfortably.

Your video never leaves the page. ffmpeg.wasm runs entirely in your browser.

About this video tool

What is this?

A free, online video trimmer and converter that runs entirely in your browser. Trim a clip, turn a video into an animated GIF, extract the audio as MP3, or compress a large file to something you can email — all without uploading the file anywhere.

Privacy — your video stays on your device

Unlike sites that "process online" by uploading to a server, this tool runs ffmpeg.wasm locally — the same ffmpeg you'd run on a server, compiled to WebAssembly. No upload, no signup, no daily limit, no watermark.

What can it do?

Trim video online — set in and out points, export the segment. Stream-copy mode preserves the original codec for a near-instant, quality-preserving cut.
MP4 to GIF — animated GIF with palette generation for true colours (not the gritty dithered look most converters produce).
Video to MP3 — extract the audio track as MP3, OGG Vorbis, or uncompressed WAV. Mono / 16 kHz options for transcription pipelines.
Compress video online — re-encode at lower quality (CRF) and / or resize the long edge to shrink huge phone recordings down to email-friendly sizes.

Supported inputs

ffmpeg reads everything you'd expect — MP4, MOV, WebM, MKV, AVI, M4V, FLV, plus audio-only files. The browser accepts uploads up to a few hundred megabytes comfortably; larger files may push the WebAssembly heap limit (a 4 GB browser cap).

Common workflows

Twitter / X clip — trim to ≤ 140 seconds, export as MP4 (stream-copy keeps the original quality).
GIF reaction — set start & end, switch to GIF mode, drop the width to 480 px, FPS to 12.
Podcast clip from a Zoom recording — extract audio as MP3 at 128 kbps, stereo, 44.1 kHz.
WhatsApp-friendly upload — compress mode, CRF 28, max long-edge 720 px.
Source for speech-to-text — extract audio as mono WAV at 16 kHz, then drop into our Speech to Text tool.

Stream-copy vs re-encode (trim)

The default Trim mode uses stream-copy — ffmpeg reads the source file, picks the chunks within the trim range, and writes them out unchanged. This is essentially instant and quality- preserving but snaps to the nearest video keyframe (typically every 1–10 seconds). Tick Re-encode for a frame-accurate cut at the cost of one more pass through the encoder.

GIF quality

Two-pass GIF encoding (default) is dramatically better than single-pass. ffmpeg generates an optimal 256-colour palette from the whole clip, then quantises every frame against it with Bayer dithering. Single-pass GIFs use a fixed palette and look noisier.

Compression — what to pick

The CRF (Constant Rate Factor) slider drives x264's quality knob. Lower = better quality + bigger file.

18–20 — visually lossless for most content
22–24 — high quality, good default for web
26–28 — fits in a WhatsApp / email attachment
30+ — aggressive, visible artifacts in motion

Resizing the long edge to 720 or 1280 px gives bigger gains than tweaking CRF alone, especially for 4K source material.

Limitations

File-size ceiling: ffmpeg.wasm allocates inside the browser's WebAssembly heap (~4 GB max). Very long 4K masters may not fit; use the desktop ffmpeg for those.
Single-threaded by default — multi-thread mode needs special HTTP headers (COOP/COEP). Most operations are still fast enough for typical clip-length material.
Web browsers can't write to arbitrary disk paths, so each export is a one-shot download.

FAQ

Is it really free?

Yes. No signup, no quota, no watermark, no API key. ffmpeg.wasm downloads once (about 30 MB) and then everything runs in your browser.

Will my video be uploaded?

No. Your file is read straight into the WebAssembly heap, processed locally, then written out as a download. The only network requests after page load are the one-time ffmpeg.wasm core download.

What's the maximum file size?

Browsers cap WebAssembly heap at around 4 GB. A 1-hour 1080p file (~4–6 GB) often won't fit; trim down to a clip first, or compress the source on a desktop. For most YouTube / phone-recording sizes this isn't a concern.

Why is trim so fast?

Stream-copy mode doesn't re-encode anything — it just rewrites the container with new boundary timestamps. The trade-off is that cuts snap to keyframes; if frame-accurate cuts matter, switch on "Re-encode".

Why is the first run slow?

The first call downloads ffmpeg-core (~30 MB) and warms up the WebAssembly runtime. Every subsequent operation reuses the same runtime and is much faster.

Can I merge clips?

Not in this version. Multi-clip concatenation is on the roadmap — let us know if it's a priority for you.

Related tools

Speech to Text — transcribe extracted audio with Whisper, also in-browser
Background Remover — for still photos exported from a video
Watermark — for image / brand overlays on stills