Transcribe Hebrew audio to text
To transcribe Hebrew audio to text, upload the file to IvreetMeet and read the transcript a few minutes later: every passage carries its speaker and a timestamp, and a summary with decisions and action items sits next to it. Audio and video up to 500MB work, including WhatsApp voice notes, and the first 60 seconds can be tried without an account.
Step by step
- 1Get the file onto your computer or phone. A Zoom or Teams recording, a voice memo, a WhatsApp voice note, or a video. MP3, WAV, M4A, AAC, OGG/OPUS, FLAC, AMR, 3GP, WMA, AIFF, CAF, and video: MP4, MOV, WEBM, MKV, AVI.
- 2Upload it. Drag it in. Add instructions if the summary should focus on something specific.
- 3Wait a few minutes. You can close the tab. The transcript appears first, the summary after it.
- 4Name the speakers. One pass, once. The names then appear throughout the transcript and in the summary.
- 5Export or share. PDF, Word, PowerPoint, or a link.
What changes the quality of the text
| Factor | Effect | What to do |
|---|---|---|
| Microphone distance | The largest single factor | Record close; use the platform recording, which captures each participant separately |
| Echo in the room | Words merge and disappear | Smaller room, soft surfaces, or a headset |
| Several people at once | The overlap is guesswork for any model | Ask people to take turns on the important parts |
| Re-compression | Removes the detail the model needs | Upload the original file, not a shrunken copy |
| Names and numbers | Rare words are the first to break | Check them in the transcript before sending the summary |
Voice notes and phone calls
A WhatsApp voice note is an OPUS file and uploads like anything else; there is a page for that case at WhatsApp voice messages to text. A recorded phone call works the same way, with one warning: recording a call you are not part of is illegal in Israel, and recording one you are part of is not. Tell the other side.
Where the text lives afterwards
The transcript stays in your organization's account until you delete it. It is not used to train models, and the audio itself is never sent to a language-model provider; only the transcript text needed for the summary is. You can delete a meeting, a speaker, or the whole account at any time.
If the result is disappointing
Frequently asked questions
Can I try it without signing up?
Yes. The home page transcribes up to 60 seconds without an account, so you can hear what your own recording sounds like as text before deciding.
What about a two-hour recording?
Long files are fine as long as they stay under 500MB. The transcript comes back in one piece, with timestamps, and stays searchable.
The audio is noisy. Is it worth uploading?
Yes, but expect more mistakes. Noise, echo and distance from the microphone cost accuracy for any speech model. If you can re-record, put the microphone closer to the speakers.
Will it keep the speakers apart?
Yes. Separation is automatic: the system marks where the speaker changes, and you give each one a name once.
Can I get the text as a file?
Yes: PDF, Word or PowerPoint through the export studio, plus a share link. A public preview link shows only the first 20%.
Try it on your own recording
IvreetMeet transcribes and summarizes Hebrew meetings. The free plan covers 7 meetings and 120 minutes per organization each month.
Start free