The container almost every meeting recording comes out of — and one that need not hold a picture at all.
You convert an MP4 to text by uploading the whole file — we read the audio track out of it, so you do not have to extract anything. Most MP4s we see are meeting recordings, and an hour of that video weighs 450 MB to 2.2 GB. The same hour as an audio-only file is 30–60 MB, so upload that one if your platform offers it.
Drop a file here or choose one from your device. We will show you the beginning of the transcript first.
skryba.ai transcribes on Scribe v2, which gets 2.2% of words wrong in Artificial Analysis' independent index — the lowest of every model measured. Gemini 3.5 Transcribe gets 2.6% wrong, Whisper large-v3 4.1%. It is the same model that writes the transcript preview you see before paying.
| Model | WER |
|---|---|
| skryba.ai (Scribe v2)the model we transcribe on | 2.2% |
| MAI-Transcribe-1.5Microsoft Azure | 2.4% |
| Gemini 3.5 TranscribeGoogle | 2.6% |
| Universal-3 ProAssemblyAI | 3.1% |
| GPT TranscribeOpenAI | 3.3% |
| Whisper large-v3OpenAI | 4.1% |
| Nova-3Deepgram | 5.2% |
Source: Artificial Analysis, AA-WER v2 index, non-streaming mode — read 27 August 2026. Seven of the 40-plus models measured: the top of the table and the names you already know. The index is measured mostly on English audio; for other languages, Polish included, no comparable public ranking exists. That is why we show you the start of every transcript before we ask for anything.
MP4 is a container, not an audio format or a video format. The same .mp4 file may carry a video track alongside the sound, or the sound alone and nothing else. An .m4a file is that identical container with a single audio track — which is where half the confusion around this format comes from.
For transcription it makes no difference whatsoever: we skip the picture and read the audio track either way. It makes a large difference to how long the upload takes.
Meetings are the most common source of the MP4s we see, and the material that benefits most from being uploaded differently. The picture in such a recording is usually a shared screen or a few near-static cameras — almost no content and a great many bytes.
The figures below are estimates and all computed the same way: 1 Mbps is about 450 MB per hour. The “fits in 5 GB” column divides the upload limit by them — the same limit for everyone, anonymous, trial and paid alike.
| What you upload | Typical source | Per hour | Fits in 5 GB |
|---|---|---|---|
| Sound only, 64 kbps mono | audio-only meeting recording | 29 MB | ~175 h |
| Sound only, 128 kbps | higher-quality export | 58 MB | ~88 h |
| Video, 1 Mbps | shared screen, little motion | 450 MB | ~11 h |
| Video, 2 Mbps | camera grid, an ordinary meeting | 900 MB | ~5.5 h |
| Video, 5 Mbps | full 1080p, plenty of motion | 2.2 GB | ~2 h |
Scroll the table sideways to see the remaining columns.
The same hour of a meeting therefore weighs fifteen to forty times less if you take the audio version. Zoom can record a separate audio-only file — it has to be switched on in the recording settings. On other platforms, look for that option where you download the recording; not all of them offer one, and then you simply upload the video.
If the meeting was recorded by a platform you have access to, its own transcript is sometimes enough — especially when all you need is a quick reminder of what was said. Uploading the same file here gives you something that transcript, by its nature, does not.
A meeting recording is usually other people's words, so it is worth remembering the participants' consent to being recorded, and who you share the finished transcript with. Recordings are stored privately, and your account settings can delete the audio file automatically after 1, 7 or 30 days — the transcript itself stays.
No. Upload the file whole — we read the audio track out of it and skip the picture. Extracting shortens the upload and nothing else, so it only pays off with a genuinely large recording or a slow connection.
Zoom can save a separate audio-only file if you switch that on in the recording settings. On other platforms, check the list of files where you download the recording. If there is no such version, upload the video — 5 GB covers several hours of a typical meeting.
No, and it is actually the easier case. MP4 is a container that can carry sound alone — an .m4a file is the same container with a single track. We take both identically and read exactly the same thing from each.
The limit is size, not length: 5 GB. For meeting video that is roughly 2 to 11 hours depending on the video bitrate. For the audio-only version it is closer to a hundred hours and beyond.
