Free online audio and video tools

Cut, convert, extract, and enhance audio and video files. Works with MP3, MP4, WAV, FLAC, MOV, and more.

Sixty-five tools cover sound and moving pictures here: format conversion for MP3, WAV, FLAC, OGG, Opus, AAC, M4A, AIFF, WMA, AC3 and APE, container work for MP4, MOV, MKV, WebM, AVI, FLV, TS, VOB and MPG, plus trimming, merging, speed and pitch changes, normalisation, equalisation, noise reduction, silence and filler-pause removal, mixing, waveform rendering, ID3 tagging, BPM and key detection, frame grabbing, thumbnails, watermarks and a full subtitle workflow. The engine behind most of them is ffmpeg compiled to WebAssembly, downloaded once into the tab and then run locally on a single thread. That single thread is the reason the limits look the way they do. A re-encode holds the input, the output and the working buffers inside a roughly 2 GB WebAssembly heap, so conversion, trimming, compression, resizing, rotation and muting accept files up to 200 MB. Two families escape that ceiling because they never decode every pixel: embedding a subtitle track is a pure stream copy, and grabbing a thumbnail or a frame uses the browser's native video decoder with a canvas, so both accept up to 500 MB. Duration guidance follows the codec cost: about 10 minutes for H.264 edits, 15 minutes for MP4, MOV and MKV conversions, and 5 minutes for WebM, which is encoded as VP8 because the VP9 encoder in the pinned core is broken. Reversing a clip is the one hard memory wall on the site — the whole decoded video has to be buffered, so the cap drops to 20 seconds at 1080p, 45 seconds at 720p and 90 seconds below that. Audio encoders available include libmp3lame, libvorbis, libopus, FLAC, AAC and uncompressed PCM at 16, 24 or 32 bit. Nothing you feed these tools is uploaded, and the output carries no watermark and no length limit beyond the ones above. One tool genuinely cannot work locally, and says so on its own page: Speech to Text drives the browser's built-in speech engine, and Chrome, Edge and Safari send the captured audio to their own vendor's transcription service. ToolChamp never receives it, but it does leave your machine. BPM and Key Detector, by contrast, decodes the track through the Web Audio API and analyses it in the page, and the subtitle set — editor, converter between SRT, VTT and ASS, timing sync, style generator, burn-in and stream-copy embed — never touches a network at all. Two recent additions are worth calling out separately. Audio Joiner exists because gluing two audio files together only looks trivial: an MP3 recorded at 44.1 kHz stereo and one at 48 kHz mono cannot be concatenated as bytes without the result playing at the wrong speed or losing a channel after the seam, so the page decodes every input first, normalises them to one sample format and joins them from there. Compress Video for Discord starts from the platform limit rather than from a quality slider: it reads the clip duration, divides the byte ceiling you pick by it to get a real bitrate budget, and encodes to that — which is the only way to stop the same CRF value producing three megabytes of a talking head and forty of gameplay.

65 tools · always free · no account needed