Get an Audio File Under 16 MB for WhatsApp

Audio file size is just bitrate × duration, so there are exactly two levers: encode at a lower bitrate, or send less audio. The arithmetic to keep in your head is that 16 MB holds roughly 17 minutes at 128 kbps, 35 minutes at 64 kbps, and 70 minutes at 32 kbps mono — and for speech, 64 kbps mono is genuinely fine, which means most "too large" voice recordings fit with room to spare once they stop being WAV files.

On Android, Convertlix's Audio Converter handles MP3, AAC, WAV, FLAC and OGG and can trim, all on the handset, so the recording never leaves the device — no upload, no server-side copy, nothing to retain. The scope, up front: Android 8.0 or later, sign-in required (there is no guest mode), 50 MB per file for audio on the free tier against a server-set monthly allowance of conversions shown in the app, and the output lands in Downloads. Everything below applies whichever tool you use.

First: check whether you need to compress at all

Before touching the audio, know this, because it solves the problem outright for a lot of people.

WhatsApp applies a much smaller cap to audio and video sent as media than to files sent as documents. As of September 2026 the media path is commonly limited to around 16 MB, while files attached through the document picker can be far larger — up to 2 GB. Both numbers can change and vary by client version, so treat them as current guidance rather than permanent facts.

So the first thing to try is not compression:

Send it as a document, not as audio. In the attachment menu, choose Document rather than Audio and pick the file.

What you give up is the in-chat player and the waveform — the recipient gets a file to tap and open in their own player, and on some clients it will not stream while downloading. For a 40-minute interview recording that is a completely reasonable trade, and it beats degrading the audio.

Compress when the recipient needs to play it in the chat, when they're on a metered or slow connection, or when you are sending the same file to a lot of people.

The arithmetic

File size in megabytes ≈ bitrate (kbps) × duration (seconds) ÷ 8,192. Turned around, for a 16 MB ceiling:

Two things fall out of that table.

A WAV file is the whole problem. Uncompressed 44.1 kHz 16-bit stereo audio runs at 1,411 kbps, which is 10 MB per minute. A three-minute WAV is around 31 MB before you have done anything wrong. If your file is a WAV, converting it to MP3 or AAC — with no other change — cuts it by roughly 90%, and for speech the difference is inaudible on a phone speaker. Same for FLAC: it preserves every original sample and is typically only 40–60% smaller than WAV, so it is still far too big for this job.

Mono halves it, and speech is mono anyway. A voice recorded on a phone has essentially the same signal in both channels. Storing it twice buys you nothing. For any spoken-word recording, mono at 64 kbps is the setting that makes this problem disappear.

Bitrate for speech vs music

The reason 64 kbps mono is fine for speech and dismal for music is bandwidth. Human speech carries almost all of its intelligibility below about 8 kHz; cymbals, string harmonics and synth textures live above that and are the first thing a low-bitrate encoder discards. So:

• Speech (interview, lecture, voice memo, meeting): 64 kbps mono is the sweet spot. 48 kbps is still clearly intelligible. Below about 32 kbps voices start to sound underwater.
• Speech with music (podcast with a theme): 96 kbps, and mono if the music is incidental.
• Music you want people to enjoy: 192 kbps stereo minimum, 256–320 kbps if you care. Do not send music at 64 kbps and expect anyone to thank you.

Trimming is the other lever and it is frequently the better one. Most recordings have two minutes of room noise at the start while someone sets up, and a tail after the useful part ends. Cutting genuine dead air costs nothing in quality, unlike every bitrate reduction. If what you actually want is the audio out of a video file, Extract the Audio From a Video on Android covers that as a separate job — extracting can often copy the existing audio stream without re-encoding at all.

How to get an audio file under 16 MB

1. Try sending it as a document first. It takes ten seconds and it may end the task without touching the file. Check that your recipient can open it.
2. Check what format it currently is. A WAV or FLAC file is the cause of the problem, not a symptom — converting it to MP3 or AAC is the single biggest saving available and it comes before any setting.
3. Trim the dead air. Cut the setup at the start and the trailing silence. This is free size reduction, and Convertlix's Audio Converter trims as well as converts.
4. Convert to MP3 or AAC, mono for speech. In Convertlix, open Audio Converter, pick the file through Android's picker and choose your output format. For a spoken recording, mono is the setting that halves the file with no perceptible cost.
5. Match the bitrate to the content, using the table above. Speech: 64 kbps. Speech with music: 96 kbps. Music: 192 kbps or more, and accept a shorter duration under the cap.
6. Check the size and listen to the worst part. Read the file size in Downloads, then play back the quietest passage or the most heavily accented speaker. That is where low-bitrate damage shows first.

When this doesn't work

• Long, high-quality music. A 45-minute set under 16 MB means about 48 kbps, and music at 48 kbps sounds bad in a way no encoder setting fixes. Send it as a document or share a link.
• Audio for transcription or archive. Do not compress the master. Transcription accuracy drops with aggressive compression, and a lossy re-encode is permanent. Send the compressed copy and keep the original.
• Anything already heavily compressed. Re-encoding a 64 kbps file to 64 kbps again adds a second generation of artefacts without saving space. Always start from the highest-quality version you have.
• Legal or medical recordings with evidentiary weight. Re-encoding changes the file. Keep and send the original by another route.
• Files over 50 MB on the free tier. The free per-job cap for audio is 50 MB, so a long uncompressed WAV master may exceed it before conversion. That is a limit — Premium raises it to 2 GB, or trim on a computer first.

FAQ

What is WhatsApp's audio file size limit?

As of September 2026, audio and video sent through the media attachment path are commonly limited to around 16 MB, while files sent through the document path can be much larger — up to 2 GB. The limits differ by client version and change over time, so check on your own device. The practical implication is that a recording too big to send as audio will often send fine as a document, at the cost of the in-chat player.

What bitrate should I use for a voice recording?

64 kbps mono is the right default for speech, and it fits about 35 minutes into 16 MB. Speech carries nearly all of its intelligibility below 8 kHz, so a low bitrate costs much less than it does on music, and a voice recorded on a phone is effectively mono already — storing two identical channels doubles the file for nothing. Drop to 48 kbps if you need more duration; below about 32 kbps voices start to sound distinctly processed.

Why is my recording so large when it's only three minutes?

Almost certainly because it is a WAV file. Uncompressed 44.1 kHz 16-bit stereo audio runs at 1,411 kbps — about 10 MB per minute — so three minutes is roughly 31 MB before anything has gone wrong. Converting it to MP3 or AAC cuts that by around 90%, and for speech played on a phone the difference is not audible.

Does Convertlix upload my recording?

No. The file is opened through Android's own file picker, converted or trimmed on the phone, and saved back to the device — nothing is sent anywhere, and the app has no storage permission, so it cannot read anything you did not hand it. Its network calls carry identity, licence, quota, configuration and payment data, never a recording. Sign in once, switch on airplane mode, and it still works.

Is there a free way to compress audio without an app?

Yes. FFmpeg is free, open source and the reference tool for this — ffmpeg -i in.wav -c:a libmp3lame -b:a 64k -ac 1 out.mp3 does exactly what this page describes in one command, and it will process a folder of files without a batch limit. Audacity is free, has a graphical interface, and is better than any phone app for trimming and levelling before you export. Both run locally. As of September 2026 browser-based audio converters generally upload your file to their servers first, which for a private interview or a medical note is worth knowing before you use one.

---

If the recording is on your phone and WhatsApp is refusing it, try sending it as a document first — and if that won't do, convert it to mono MP3 and check the size in Downloads.