Getting a voice memo off your phone
On iOS, open the Voice Memos app, find the recording, tap share and choose Files. The memo saves with an M4A extension and can be uploaded straight from your computer or from the browser on the device you are reading it on.
On Android, recordings usually land in a Files or Voice Recorder folder as M4A or MP3 depending on the app. Any of those upload directly.
A short M4A from a phone is well under the 100 MB limit, and phone memos are almost always inside the 10-minute limit too, so this is the one transcription case that usually needs no splitting at all.
Memos are rough, and that is fine
A voice memo is a thinking tool, not a recording. The result is full of half-started sentences, asides, and the sentence you changed your mind about halfway through. A verbatim transcript of that is not especially useful, and the instinct to preserve it exactly is usually wrong.
What you want is extraction. Read the transcript, keep the decisions and the list items, and discard the rest. This is the one transcription job where doing less with the output is the correct approach.
Memos recorded while walking or driving are the hardest case. Road noise and wind are broad-spectrum and the model will produce confident wrong words rather than obvious errors, so read anything that looks specific and verify it.
What people transcribe memos for
Lists spoken while driving or cooking are the most common use. A memo that reads back as a clean bulleted list is immediately actionable, and doing it later from the recording takes ten times as long.
Decisions and reminders are the second. "Remember to email the contractor about Thursday" becomes text you can search, and a decision made out loud is easier to revisit than one you are relying on memory to recall.
Rough drafts are the third. Speaking a paragraph of prose and transcribing it produces a first draft faster than typing, and it bypasses the blank page. Expect to edit it, since spoken drafts are full of spoken filler.
Recording other people
Recording a conversation on a phone and transcribing it raises a consent question that recording alone does not. Depending on where you are, you may be required to tell people they are being recorded, and in some places you need permission.
The safer default is to tell people. It costs one sentence at the start of the call and removes a real problem. Note also that transcripts are easier to forward accidentally than recordings, so a memo of a conversation is more widely shared than you might expect.
Konthora holds uploads and results for 60 minutes and then deletes them, which limits how long a sensitive recording sits around. That is a useful property, but it is not a substitute for telling people.
Step by step
Get the memo off your phone as a file
On iOS, share the memo to Files and it saves as M4A. On Android it is usually already in your Files app as M4A or MP3. M4A, MP3, WAV and AAC are all accepted.
Listen once and note what you want out of it
Knowing whether you want the list, the decision or the whole thing changes what you keep. Voice memos rarely survive a verbatim transcript, and that is fine — the useful part is usually a tenth of it.
Upload and choose sentence timestamps
Sentence mode is right for a memo. Memos are short and unstructured, so per-sentence timings are more than enough granularity and the output reads cleanly.
Transcribe, then cut it down
Read the transcript and keep what still matters. A memo recorded three weeks ago is a record of what you were thinking, and you now know which parts of that still apply.
Export TXT and export it early
Memos get transcribed and then forgotten. Export the text straight away rather than treating the recording as the artefact — a transcript you can search is worth more than a recording you have to replay.