Does Hushscript join my Zoom meeting as a bot?

No. Hushscript is upload-based, not a meeting integration. You record the meeting yourself, locally or to Zoom's cloud, then upload the saved file. Nothing connects to your Zoom account and no extra attendee appears in the call.

How do I get the recording off Zoom?

For local recordings, Zoom saves them to your computer automatically, usually under Documents/Zoom. For cloud recordings, log in to your Zoom account, open Recordings, and download the MP4. Either way you end up with a file you can upload.

Does the whole video file upload?

No. The audio is extracted from the video in your browser before anything is sent. The MP4 stays on your device and only the audio reaches the server, where it is deleted the moment the transcript is ready.

Can I get a transcript that shows who said what?

Yes. Speaker labels are included on every transcript automatically. After transcription you click any label to replace Speaker A with a participant's real name, and that name updates everywhere in the transcript.

What if the meeting had more than two people?

Hushscript separates multiple speakers automatically with no speaker cap. Accuracy is best when people take turns rather than talking over each other, which is common once a meeting gets lively.

Do I need to tell people the meeting is being transcribed?

Recording consent rules vary by country and state, so check the law where your participants are. Practically, you already recorded the meeting in Zoom, so the consent question is about the recording itself. Transcribing the saved file afterward adds no new attendee to the call.

How to Transcribe a Zoom Meeting or Recording

Q: Are the transcripts kept private?

The audio is deleted as soon as the transcript is ready, and the transcript itself is encrypted at rest. If the storage were ever leaked, the contents would be unreadable ciphertext rather than your meeting in plain text. You can also delete any transcript in one click.

June 10, 2026

There are two ways to turn a Zoom meeting into text. One is to let an AI notetaker bot join the call as a visible attendee and listen in real time. The other is to record the meeting yourself and upload the saved file afterward. Hushscript is the second kind. It is not a Zoom integration, a meeting bot, or a notetaker that sits in your call.

That difference is the whole point of this guide. With the upload approach, no third-party service is present while you are talking, nothing is granted access to your Zoom account, and no extra name shows up in the participant list for everyone to see. You keep control of the recording from the moment it is saved.

Below is the full process: getting the recording off Zoom, extracting the audio, producing a speaker-labeled transcript, and keeping the whole thing private. It works for a two-person one-on-one and for a board meeting with eight people on the call.

What you need

You need three things, and you almost certainly already have all of them.

The first is a recording of the meeting. Zoom can produce one for you, either saved to your computer or to your Zoom cloud account, and the next section covers both. The second is the right to transcribe it, which usually means you were a participant and recorded it yourself. The third is a browser. Hushscript runs in Chrome, Firefox, Safari, or Edge, and there is nothing to install.

You do not need a Zoom paid plan for local recording, and you do not need to connect any apps or grant calendar access. The file does the work.

Record the meeting, local or cloud

Zoom gives the host two recording modes, and which one you pick decides where the file ends up.

A local recording saves straight to the computer of whoever clicked record. You control the file, and there is no cloud copy unless you make one. By default Zoom stores local recordings under Documents/Zoom/, in a folder named for the date and meeting title. Inside you will usually find an MP4 video and sometimes a separate M4A audio file. Local recording is free on every Zoom plan and is the simplest path if you are the host.

A cloud recording uploads to your Zoom account instead. To get the file, log in at zoom.us, open Recordings, find the meeting, and download the MP4. Cloud recording requires a paid Zoom plan and is the host’s by default. Other participants can only reach it if the host shares a link, which is worth remembering when you think about who has a copy.

Either route leaves you with a file on your computer. That file is what you upload, and Hushscript does not care whether it came from local or cloud recording.

Transcribe the recording in three steps

The flow is the same whether your file is the full MP4 or a separate audio track.

Drop the file and watch the preview. Go to audio to text and drop your MP4 or M4A onto the upload area. The audio is pulled from the video in your browser, and within a few seconds you get a 30-second speaker-labeled preview of the start of the meeting. No account is needed for this step, so you can check the labeling and accuracy before committing to anything.
Sign up to transcribe the rest. If the preview looks right, create an account with your email. This is the gated step. The 30-second taste is open to everyone, but transcribing the full meeting needs an account. You get 30 free minutes to try the real thing, granted the moment you validate a card with a $1 hold that is released right away and never charged, or with your first purchase if you pay another way. A card is not required.
Upload the full file and get the transcript. Back in the app, upload the whole recording. The transcript comes back with one utterance per line, each attributed to a speaker label such as Speaker A or Speaker B, with timestamps. From there you relabel the speakers and export.

Only the audio ever leaves your device. The MP4 stays put, the extracted audio is deleted the instant the transcript is ready, and what remains is text you own.

A worked example

Say you hosted a 45-minute quarterly review with four people and recorded it locally. Zoom dropped an MP4 into Documents/Zoom/2026-06-18 Q2 Review/. You drop that MP4 onto the upload area, the first 30 seconds preview cleanly with two distinct speakers already separated, and you sign up to do the rest.

A few minutes after uploading the full file, you have a transcript that looks like this:

Speaker A [00:05:22]: The Q2 numbers are in. Margins held but volume is down twelve percent.
Speaker B [00:05:31]: That matches what procurement flagged last week.
Speaker A [00:05:38]: Right. So the question is whether we adjust the forecast now or wait for July.
Speaker C [00:06:02]: I would wait. One soft month is not a trend yet.

You click Speaker A, type Sarah, and every line she spoke is relabeled at once. You do the same for the others. Five minutes of clicking turns four anonymous voices into a named record of who said what. You export it as DOCX, circulate it to the four attendees, and the meeting is documented without anyone having to write notes during the call.

Speaker labels for who said what

A flat wall of text is hard to act on. A meeting transcript that shows who said each line is something you can actually use, and Hushscript adds those labels to every transcript by default, with no speaker cap and no paid tier.

Putting real names on the labels takes three clicks. Click any instance of Speaker A in the editor, type the person’s name, and the label updates everywhere that speaker appears. For a 60-minute meeting with five people, the relabeling pass takes a minute or two. For a 30-minute one-on-one it takes under a minute.

Once the names are in, the transcript reads like the conversation it was. You can scan for a single person’s contributions, pull every commitment one participant made, or quote a decision with the right name attached. For more on how the speaker separation works under the hood, see speaker identification.

Keep the meeting private

Internal meetings carry the things you least want leaking: budgets, headcount, strategy, legal exposure, candid opinions about people who are not in the room. How the meeting becomes text matters as much as the text itself.

When a notetaker bot joins a call, it is a live participant. It is visible to everyone, it records from inside the meeting session, and your conversation streams through that service’s infrastructure while you are still speaking. Some participants will clam up the moment they see an unfamiliar attendee labeled as a recorder, and you cannot always control where that real-time stream goes.

Upload-based transcription removes the bot from the equation. The recording already exists on your device before Hushscript is ever involved. The audio is extracted locally in your browser, so the video file never uploads. The audio that does reach the server is deleted the moment the transcript is ready, and nothing ever touches your Zoom account, your calendar, or your participants.

There is one more layer worth knowing about. Your transcripts are encrypted at rest. If our storage were ever leaked, the contents would be unreadable ciphertext rather than your meeting in plain text. That is leak protection, not a claim that no one on our side can read it: the key is held on our servers, not by you alone. The fuller picture of how the data is handled lives on the private transcription page.

Troubleshooting common Zoom recordings

Real meeting audio is messier than a clean studio recording, so here are the issues that come up most and how to handle them.

People talk over each other. Quarterly reviews and brainstorms get heated, and when two voices overlap the system has to guess where one speaker ends and the next begins. Crosstalk is the single biggest cause of mislabeled lines. You cannot fix the audio after the fact, but you can fix the transcript: scan the busy stretches, and where a label looks wrong, correct it. Going forward, asking people to mute when not speaking does more for transcript quality than any setting.

One person is much quieter than the rest. If a participant dialed in from a laptop mic across a room while everyone else used a headset, their lines may be weaker or occasionally dropped. Zoom’s own audio settings help here. Turning on “Optimize for original sound” before recording, and asking remote attendees to use a headset, lifts the quiet voice closer to the others.

The file is large. A long video meeting can be a sizeable MP4, but the size that matters is the audio, not the video. Because Hushscript extracts the audio in your browser before upload, a multi-gigabyte MP4 becomes a much smaller audio file on the way out. Files up to 10 hours and 2 GB are supported, which covers any normal meeting, and there are no daily limits beyond the minutes you have.

Strong accents or technical jargon. A room of people from different countries, or a call full of product code names, can trip up any transcriber on the odd word. The model handles a wide range of accents well and detects the language automatically across roughly 99 languages, listed on the languages page. For the rare misheard term, a quick find-and-replace in the exported file fixes every instance at once.

You are not sure who Speaker A is. When the meeting opens with small talk, the first labeled voice is not always the host. Use the timestamps to jump to a moment where someone introduces a topic you remember, and identify them from context. Once you have matched one or two voices, the rest usually fall into place quickly.

The recording is only audio. If Zoom produced an M4A instead of an MP4, nothing changes. Drop the M4A in directly. There is no video to extract from, so it goes straight to transcription.

Local vs cloud recording: which to transcribe

If both a local and a cloud copy exist, transcribe whichever is more convenient, because the transcript is identical either way. The choice matters more for control than for quality.

A local recording lives only on your machine until you decide otherwise, which is the better fit for a sensitive meeting where you want the fewest possible copies. A cloud recording is easier to retrieve from another device and harder to lose if your laptop dies, at the cost of an extra copy sitting in Zoom’s cloud. For routine team syncs, cloud is convenient. For anything confidential, a local recording keeps the file in your hands from the start.

Whichever you upload, the privacy of the transcription step is the same: audio in, transcript out, audio deleted.

Tips for a clean meeting transcript

A little care before and after recording goes a long way.

Before the meeting, encourage headsets over laptop microphones and ask people to mute when they are not speaking. Recording at a decent audio quality, rather than the lowest bandwidth setting, gives the model more to work with. None of this requires a studio, just slightly better habits than the average call.

After transcription, do the speaker relabeling first while the meeting is fresh in your memory, then read the transcript once against your recollection of the busy moments. A single pass catches the handful of crosstalk slips, and the result is a record you can circulate with confidence.

Hushscript runs on the same top-tier AI for a meeting as for a studio interview, so it is not free. It is pay-as-you-go with no subscription, and you pay only for the minutes you transcribe, as little as about a penny a minute on the largest pack. The current pack prices are on the pricing page.

Frequently asked questions

The questions above cover the practical points. For the same speaker-labeling workflow applied to a two-person conversation, see how to transcribe an interview, and for the multi-speaker show-notes angle, how to transcribe a podcast. The approach for Microsoft Teams and Google Meet is the same as for Zoom: record the call, save the file, and upload it to audio to text.