Saidby
Private Meeting Transcriber: Who Said What
Saidby is a private meeting transcriber for iPhone, iPad and Mac. Record a meeting, or drop in a recording you already have, and get back a transcript that says who said what, with a name and a time on every paragraph. Transcription and speaker separation run entirely on your device, with no networking code, no account and nothing ever uploaded, and you buy it once.
Key Features
- Who said what: tell it how many people were in the meeting (up to 20) and Saidby separates and labels the voices itself, introduces each one by its first sentence with a play button so naming it takes seconds, and merges a voice split in two with a tap
- Record in the app with the screen locked, in the background and through a phone call, from the built-in or a connected microphone, or import voice memos, video-call recordings, MP3, M4A, WAV, MP4 and MOV
- Choose each meeting's language from those your device transcribes offline, including English, French, German, Spanish, Italian, Portuguese, Japanese, Korean and Chinese, then tap any timestamp to hear that moment, edit lines, reassign paragraphs and export Markdown, plain text, SRT or JSON
- Private by construction: Apple's on-device speech models, no networking code, no account, no analytics and a Data Not Collected privacy label, for one price with no subscription, no in-app purchases and no minute limits
About Saidby
Saidby is a private meeting transcriber for iPhone, iPad and Mac. Record a meeting, or drop in a recording you already have, and you get back a transcript that says who said what, with a name and a time on every paragraph. The whole job happens on your device, and nothing is uploaded, ever — not as a setting you have to remember to switch off, but because there is nothing in the app that could send it anywhere.
When the meeting ends, Saidby does three things: it separates the voices, writes down the words, and lines the two up. You tell it how many people were in the room, up to 20, and it separates their voices by itself and labels each one. Speaker changes land on sentence boundaries rather than in the middle of a word. Each voice is then introduced by its opening sentence with a play button beside it, so putting a name on it takes seconds — and once a voice has a name, the whole transcript updates. If one person has been split into two voices by mistake, you merge them with a tap.
Getting audio in is flexible. You can record in the app with the screen locked and the phone in your pocket: recording keeps going in the background and survives a phone call, and you choose the input, the built-in microphone or a connected one. Or you can import a file you already have — voice memos, video-call recordings, MP3, M4A, WAV, MP4 and MOV all work.
Each meeting has its own spoken language, chosen from the ones your device transcribes offline: English, French, German, Spanish, Italian, Portuguese, Japanese, Korean, Chinese and more. Typography follows the language, so a French transcript gets its spacing before punctuation and proper quotation marks, with no stray spaces.
The result is a transcript you can work with rather than a wall of text. Every paragraph carries a timestamp; tap one to hear that moment again. You can edit any line, move a paragraph to another speaker, and rename anyone at any time. When it is ready, export it as Markdown, plain text, SRT subtitles or JSON, through the share sheet or straight to Files. Keep the audio next to the transcript, or delete it once the notes are written.
Who is Saidby for?
- Consultants and managers who record client meetings but cannot upload a client’s voice to somebody else’s server
- Journalists and researchers with interviews to transcribe and keep, including interviews under embargo
- Students with lectures they would rather reread than scrub through
- Anyone recording sensitive conversations, such as client meetings or medical consultations, where the audio should never leave the device
- People with a video-call recording or a voice memo who want a readable transcript with names on it, not just words
- Anyone who has ever come out of a meeting asking “who said that?”
Saidby is built for conversations between several people, and it transcribes once the recording ends. To dictate notes on the go and watch the words appear as you speak, Transcribe, our speech-to-text app for iPhone, is the better fit.
Why choose Saidby?
Most transcription services work by sending your audio to a server, and for a lot of conversations that is a non-starter: a client’s voice, a medical consultation, an interview under embargo. Saidby takes the other route. Transcription uses Apple’s on-device speech models, and speaker separation runs inside the app. Saidby has no networking code at all — no server, no cloud, no upload, no sync — and no account, no sign-in, no analytics and no advertising identifier. Its App Store privacy label reads “Data Not Collected”, which is a fact about how the app is built rather than a promise. The recording and the transcript stay in Saidby’s own folder on your device, and a transcript reaches another app only when you export or share it yourself.
It also works offline. The system downloads the speech model for each language once, and from then on transcription needs no connection at all, wherever the meeting happened.
And it is a single fair price. Buy Saidby once and the whole app is yours: no subscription, no in-app purchases and no minute limits. Universal purchase covers iPhone, iPad and Mac, and Family Sharing is enabled, so one purchase serves every device and everyone in the household. Saidby requires iOS 26 or macOS 26 or later, and it is made in Paris. Record the meeting; read who said what.