The shortlist
6 things that do it.
The built-in iPhone recorder, which can show a transcript while you record or afterwards.
- Free
- No separate price is stated on the Apple page; the app is part of iOS.
- Paid
- price not checked
- Best at
- Transcribing speech on the phone you already have, with the ability to view, copy and search the transcript.
- The catch
- Apple says audio transcription needs iPhone 12 or later, covers English, Spanish, Portuguese, Italian, French, German, Japanese, Korean and Chinese, and is not available in all countries or regions.
- ios
The recorder app on Google Pixel phones, with transcription of what it records.
- Free
- No separate price is stated on the Google page; the app comes with Pixel phones.
- Paid
- price not checked
- Best at
- Transcribing what it records, built into Pixel phones.
- The catch
- It is for Pixel phones (Pixel 3 and later per Google). Re-transcribe may process audio on Google servers, and Google says some languages may need internet access for summaries.
- android
Transcribe in Word and OneNote for Microsoft 365: record or upload audio and get a timestamped transcript.
- Free
- No free tier; included with a Microsoft 365 subscription (300 minutes of uploads a month).
- Paid
- price not checked
- Best at
- Turning an uploaded .wav, .mp4, .m4a or .mp3 file into a Word-editable transcript in 80+ locales.
- The catch
- Needs a Microsoft 365 subscription, an internet connection, and is capped at 300 minutes of uploaded audio a month on a Microsoft 365 plan.
- windows
- web
AI meeting and conversation transcriber with apps and a Chrome extension.
- Free
- Basic: 300 monthly transcription minutes, up to 30 minutes per conversation, 3 lifetime file imports (Otter pricing page).
- Paid
- from $8.33 per user per month (Pro, billed annually)
- Best at
- Live transcription of meetings with speaker recognition, plus summaries, on web, desktop and mobile.
- The catch
- The free Basic plan has 300 monthly transcription minutes, 30 minutes per conversation and only 3 lifetime file imports; Pro allows 10 monthly imports.
- web
- ios
- android
- mac
- windows
- chrome
Open-source speech recognition model from OpenAI, released under the MIT licence.
- Free
- Free: code and model weights are MIT licensed.
- Paid
- from Free (MIT licence)
- Best at
- Free, local transcription for people comfortable with the command line, in many languages.
- The catch
- You install it with pip and need FFmpeg (and sometimes Rust); OpenAI says performance varies significantly by language.
- linux
- mac
- windows
- api
A Mac app that transcribes files, meetings and system audio using Whisper and other models.
- Free
- Free version: basic transcription, recording and dictation (MacWhisper site).
- Paid
- from €64 one-time (Pro)
- Best at
- Private transcription on a Mac: the site says it uses local models so files do not leave the Mac.
- The catch
- Mac only. Speaker recognition and batch transcription are Pro features.
- mac
The same shortlist, as a table
| Thing | Free tier | Paid from | Platforms | The catch |
|---|---|---|---|---|
| Apple Voice Memosrel="nofollow noopener" | No separate price is stated on the Apple page; the app is part of iOS. | not checked | ios | Apple says audio transcription needs iPhone 12 or later, covers English, Spanish, Portuguese, Italian, French, German, Japanese, Korean and Chinese, and is not available in all countries or regions. |
| Google Recorder (Pixel)rel="nofollow noopener" | No separate price is stated on the Google page; the app comes with Pixel phones. | not checked | android | It is for Pixel phones (Pixel 3 and later per Google). Re-transcribe may process audio on Google servers, and Google says some languages may need internet access for summaries. |
| Microsoft Word Transcriberel="nofollow noopener" | No free tier; included with a Microsoft 365 subscription (300 minutes of uploads a month). | not checked | windows, web | Needs a Microsoft 365 subscription, an internet connection, and is capped at 300 minutes of uploaded audio a month on a Microsoft 365 plan. |
| Otterrel="nofollow noopener" | Basic: 300 monthly transcription minutes, up to 30 minutes per conversation, 3 lifetime file imports (Otter pricing page). | $8.33 per user per month (Pro, billed annually) | web, ios, android, mac, windows, chrome | The free Basic plan has 300 monthly transcription minutes, 30 minutes per conversation and only 3 lifetime file imports; Pro allows 10 monthly imports. |
| OpenAI Whisperrel="nofollow noopener" | Free: code and model weights are MIT licensed. | Free (MIT licence) | linux, mac, windows, api | You install it with pip and need FFmpeg (and sometimes Rust); OpenAI says performance varies significantly by language. |
| MacWhisperrel="nofollow noopener" | Free version: basic transcription, recording and dictation (MacWhisper site). | €64 one-time (Pro) | mac | Mac only. Speaker recognition and batch transcription are Pro features. |
How we checked
What we read, and what we haven't tested yet.
On 2026-10-02 we read Apple's iPhone User Guide page on Voice Memos transcription, Google's Pixel support page on Recorder transcription, Microsoft's Transcribe in Word support page, Otter's pricing page and homepage, OpenAI's Whisper GitHub page and the MacWhisper homepage. Nothing was tested hands-on, so this is what each company says about its own product.
If none of these fit
The manual way.
If your phone or computer isn't covered, Whisper is free open-source code you can run on your own machine (it needs pip and FFmpeg), and MacWhisper puts a Mac app on top of the same kind of models. Apple lists the ten languages Voice Memos covers, so check yours before relying on it. None of our products transcribes audio.