Select your file
An MP4 clip, MP3 track or WAV recording. Start with the original when you have it.
Local file access
The browser reads the file you select. This step does not send its contents to a conversion server.
Compress MP4, MP3, WAV, M4A, and many other formats without uploading your media. Make smaller files for sharing or prepare audio for transcription.
Choose one file. It stays on your device.
One file at a time
First use downloads the compression engine. Your selected media is not uploaded. How privacy works
Select a file, choose what matters most and save a new copy. There is no media upload queue in between. Open a step below to see what happens inside the browser.
An MP4 clip, MP3 track or WAV recording. Start with the original when you have it.
The browser reads the file you select. This step does not send its contents to a conversion server.
Balanced for everyday use. Stronger compression for size. Speech-to-text for voice.
Keep video for anything viewers need to see. Choose the speech preset when only the spoken recording matters.
The service re-encodes the media, using your device's resources.
The processing code downloads on first use. Encoding speed depends on the file, settings, processor and available memory.
Download the new copy. Compare its size and play a section before sending it on.
Check quiet speech, music detail or small text in a video. If quality drops too far, return to the original and use Balanced.
Server-based compressors need your media before they can process it. This tool takes a different approach: the file stays with you, and your browser does the encoding.
Compress interviews, personal clips and work recordings without sending the selected media to a remote conversion service. Local processing is the default, not an extra setting.
Need the conversation, not the picture? The speech-to-text preset removes video and creates 16 kHz mono audio. Use it when your transcription service accepts those settings.
Choose a file and get started. Compression is free, with no account required and no branding added to the output.
The tool creates a new downloadable file rather than overwriting the source. Keep the original for editing, archiving or a higher-quality export later.
The format is only part of the story. A file extension names its container, while the codec describes how the media inside is encoded. Common inputs are listed below; encrypted files and unsupported codecs will not process.
| Source file | Recommended output | Use it for |
|---|---|---|
| Video: MP4, MOV, MKV, AVI, WebM | MP4 with H.264 video and AAC audio | Video clips, screen recordings and everyday web playback. |
| Audio: MP3, M4A, AAC, OGG, FLAC | MP3 or M4A with AAC audio | Smaller copies of voice notes, podcasts and music. |
| WAV recordings | MP3 or M4A | Sharing a large recording. Keep the WAV source for editing. |
| Speech in an audio or video file | Audio-only MP3 or M4A, using the 16 kHz mono speech preset | Preparing a spoken recording for a service that accepts these settings. |
Keeping the picture? Choose MP4. Selecting MP3 or M4A for a video saves an audio-only copy, not a smaller playable video.
An uncompressed PCM WAV stores far more audio data than a typical MP3. Choose MP3 or M4A to make a smaller sharing copy. Keeping WAV does not automatically deliver a large reduction, and renaming “.wav” to “.mp3” changes nothing about the data inside.
These formats already use compression. Start with Balanced, then try stronger compression if the file is still too large. Listen to quiet speech, cymbals and background detail before replacing anything. For music, skip the speech preset: mono, 16 kHz audio is not a music export setting.
Compression is not a magic cleanup button. Lossy formats remove detail to save space, and repeated exports remove more. Work from the original, use the least aggressive setting that meets your needs, and check the result.
For audio, bitrate is a useful starting point: it is the amount of data stored for each second of sound. Video size also depends on motion, resolution and the encoder's quality settings, so a fixed percentage reduction tells you very little without knowing the source.
A voice note, a screen recording and a music track call for different choices. Match the output to the job, not just the smallest number.
Choose MP3 or M4A for a spoken recording. Check the app's current attachment limit, then compare the downloaded file's size. A preset is a starting point, not a guarantee that every recording will fit.
Export to MP4 and start with Balanced. For a screen recording, check that small text is still readable. If the clip remains too large, trim unnecessary footage in an editor rather than sacrificing every frame's clarity.
An interview does not need its video track to become a transcript. Use the speech preset for an audio-only copy. Keep the original when separate left and right channels help identify speakers, and check your transcription service's input requirements.
The compressor does not upload your selected media file. It processes that file in your browser. The webpage still connects to the internet to load the engine and other resources, including analytics and advertising measurement.
The speech preset is not an automatic handoff to a transcription service. If you later choose SpeechText.AI transcription, uploading the recording is a separate step.
Save the result before closing the tab. For regulated or confidential material, follow your organization's handling rules; local processing is not a compliance certification.
Provided by SpeechText.AI. The guidance on this page explains format choices, file-size calculations and the limits of local processing.
Quality, compatibility and file limits matter more than a headline reduction percentage. Here is what to check.
Yes. You can compress files without creating an account or paying for compression. The tool does not add a watermark. SpeechText.AI transcription is a separate service.
No. The compressor processes the selected media file in your browser. The page still downloads its processing engine and website resources and uses analytics and advertising measurement. Those requests are separate from your media file.
Lossy compression removes detail to save space. Start with Balanced and compare a short section with the original. Already compressed MP3, M4A or MP4 files offer less room for reduction, and a new export can be larger than the source.
For an uncompressed WAV recording, choose MP3 or M4A to create a smaller sharing copy. For MP3, start with Balanced and try stronger compression only if the result is still too large. Renaming the file extension does not compress audio.
Yes. Choose MP3 or M4A for audio and MP4 if you need to keep the video. Check the receiving app's current size limit and the downloaded file's size before sending. This tool does not target an exact attachment size.
Available device memory sets the practical limit. Start with files below 250 MB as a guideline, not a guaranteed maximum. Long videos need extra working memory. For large recordings, use a desktop browser, close other heavy tabs or trim the source first.
No. It removes video, mixes the audio to mono and resamples it to 16 kHz. It prepares an audio file; it does not transcribe, remove background noise or repair unclear speech. Check the transcription service's accepted formats and keep the original if separate channels matter.
This interface offers presets, not manual width, height, frame-rate or target-size controls. Use it for a quick local export. HandBrake or another desktop encoder is a better choice for batch jobs and detailed video settings.
Use a current desktop browser and allow the engine download to finish. Blocked engine files, unsupported codecs and low memory are common causes of failure. Try a short, known-good MP4 or MP3 to separate a file problem from a browser problem.
Start with Balanced. Switch to stronger compression only when you need to save more space.