Transcription App — 100% private AI transcription
Transcription App turns your audio and video recordings into accurate, editable transcripts without sending a single file to the cloud. It runs OpenAI’s Whisper models and NVIDIA’s Parakeet model directly on your machine, recognises who is speaking, and gives you a complete workspace to correct, subtitle and export your work. Transcriptions are serverless: your recordings never leave your computer.
Whether you are a journalist transcribing interviews, a qualitative researcher transcribing hours of fieldwork, a Film & TV professional creating and styling subtitles, or you need a careful tool for legal audio transcription, Transcription App was built for the way you work.
Transcription App is free forever with Whisper Base. Transcription App is freemium: download it, transcribe complete recordings of any length in as many projects as you like, and pay nothing. A subscription unlocks the more advanced models when you need extra accuracy or speed.
Download on MacOS
Download on Windows
Download on Linux
AI-Powered transcription
Choose between five on-device models, from the NVIDIA’s fast Parakeet v3 to Whisper Large Turbo, which delivers accuracy close to Whisper Large v3 in a fraction of the time. Pyannote’s speaker models then split the transcript into speaker turns. Whisper Base is free forever; the other models come with a subscription.
100% Privacy-Oriented
Your recordings are processed on your device and stored in a local folder. Nothing is uploaded for transcription, ever. The app only goes online to check your licence and to download a model the first time you use it.
Edit and work on your transcripts
A calm, fast editor built around your recording. Click a line to hear it, fix a word, rename a speaker or merge two segments, then export your work as a Word document, plain text, SRT or VTT subtitles, or structured JSON.
Try the look and feel of Transcription App
This is what Transcription App does
The transcription took 8 seconds on a Macbook M1 Pro with 16Gb of RAM on default settings. Speaker recognition was done at 1% threshold minimum similarity, 0% confidence margin. The result is unedited.
The player on this page is a live, read-only replica of the Transcription App editor. It shows the real output of four of our models on the same recording, so you can compare accuracy and speed for yourself.
We have not changed a single word of these transcripts. What you read is exactly what each model produced. Here is what you can do with the demo:
- Switch models with the tabs and compare the transcripts line by line.
- Click any segment to play the audio from that exact moment.
- Right-click a segment, or Ctrl/Cmd-click it, to see its timestamps and speaker.
- Turn on Text view to read the transcript as flowing paragraphs grouped by speaker.
- Filter by speaker to see only the passages you care about.
In the app itself, the same screen also lets you:
- Edit the text directly, with undo and redo for your recent changes
- Split and merge segments, and adjust their timestamps
- Rename a speaker once and update every matching segment
- Navigate by sound with a waveform under the transcript
- Keep your project notes open beside the transcript
- Export to .doc, .txt, .srt, .vtt or .json
From recording to finished transcript in five steps
Transcription App is organised around projects, so a single interview series, research study or film production keeps its recordings, transcripts and notes in one place. Getting from raw audio to a clean transcript takes a few clicks.
- Create a project for a client, a study or a production.
- Add your audio or video files. Set the language, and switch on word-level timestamps and speaker segmentation if you need them.
- Queue as many files as you like. The app transcribes them one after another while you keep working on something else.
- Pick a model. Whisper Large Turbo is our recommended default, and lighter models are there for more modest computers.
- Review, correct and export. Your edits are saved automatically as you type.
Video files are transcribed from their audio track, and very large recordings are converted automatically by a media engine bundled with the app, so there is nothing else to install. Hours-long interviews, full episodes and multi-camera rushes are all fair game.
Five models, and one that suits your computer
Every computer is different, and so is every recording. That is why Transcription App lets you choose the model that gives you the best trade-off between speed and accuracy. All of them run entirely on your device.
- Whisper Large v3 Turbo: accuracy close to Whisper Large v3 with much faster processing. This is where we recommend you start.
- Parakeet TDT 0.6B v3: NVIDIA’s fast speech model, well suited to long recordings and fast transcripts. Set the language of your file for the best results, as automatic language detection can reduce accuracy.
- Whisper Medium: generally more accurate than Small and Base, at the cost of some speed.
- Whisper Small: a balanced option for laptops with less memory.
- Whisper Base: the fastest Whisper model, handy for quick drafts and older machines.
Whisper Base is free forever, with no limit on recording length. The four more advanced models come with a subscription.
On Macs with Apple silicon, the app uses the graphics processor by default. On other computers it starts on the main processor and can use the graphics card where it is supported. If a combination does not work on your machine, the app falls back automatically and keeps your queue moving. Advanced users can pick the compute device, the number of processor cores and the model precision, while everyone else can simply leave these settings on automatic.
Transcription App is truly cross-platform! It is available for macOS, on both Apple silicon and Intel, as well as for Windows and Linux on x64 and arm.
Know who said what
A transcript is only useful if you can tell the voices apart. Transcription App gives you two complementary ways to identify speakers, both powered by Pyannote models that run on your computer.
Automatic speaker segmentation and recognition
Use our speaker recognition functionalities to automatically populate speaker labels. Label a few segments for each person, ideally three or more, then run Speaker Recognition. You can tune the segmentation precision to match the recording, whether it is a two-person interview or a lively panel discussion.
Tidy up in seconds
Rename « Speaker 1 » to « Interviewer » once and every matching segment follows. Each speaker keeps the same colour throughout the project, so you can scan a long transcript at a glance. Merge consecutive segments from the same person into a single block, or switch to Text view to read the conversation as clean paragraphs. When you export to Word or plain text from Text view, those paragraphs are exactly what you get.
For qualitative research: meet Qualitativist
Transcription App is the focused tool for transcription, speakers and subtitles, and it covers the first step of any qualitative study: turning interviews, focus groups and fieldwork into clean transcripts without moving participant data off your computer.
Notes and search
Write memos in project notes that open in their own window, and keep them visible beside any transcript while you work. Export them as text files, or as Word files with a subscription. Keyword search finds a word or a phrase inside a transcript or across a whole project, and you can filter the results by speaker.
Need to code and analyse your data? Use Qualitativist
Our sister product, Qualitativist, is a complete qualitative research suite built on the same private, on-device transcription. It adds colour-coded tags for coding, advanced search with AND and OR rules, a literature library with citations in Word, research documents and web data collection, all kept on your own computer.
Subtitles, styled cue by cue
For Film & TV professionals, Transcription App doubles as a subtitle editor. Turn on subtitle editor mode and every segment becomes a subtitle you can preview and style directly on top of your video.
- Preview subtitles on your video, including in full screen, exactly as viewers will see them.
- Style each subtitle individually: position, width, font, text size, line spacing, text and background colours, outline, shadow, alignment, characters per line and timing offset.
- Save a look as a preset for the whole project, then apply it to one subtitle, to every subtitle you have not styled yet, or to all of them.
- Step through subtitles with the keyboard, using a shortcut you can choose yourself.
- Export to SRT for maximum compatibility, free, or to WebVTT with per-subtitle positioning and styling with a subscription.
Every change can be undone, and splitting a subtitle keeps its formatting. Note that external video players differ in how much WebVTT styling they display.
Pricing
Start for free: Transcription App is freemium. The free mode is free forever, with no time limit and no cap on recording length, and runs on Whisper Base, so you can transcribe real work without paying anything. When you want more accuracy or speed, a subscription unlocks the advanced models.
Monthly
$2.29
Yearly
$22.99
Prices are also available in your local currency!
Please contact sam@transcription-app.com for discounts on multiple licenses or if you’re representing an organization.
Free forever, with Whisper Base:
- Free forever, not a trial: transcribe complete recordings of any length with Whisper Base.
- Unlimited projects, the transcription queue, speaker segmentation and recognition.
- The full transcript and subtitle editors, notes and keyword search.
- Exports to .txt, .srt and .json, transcript imports and full-storage backups.
- Your work stays yours: everything remains accessible if a subscription ends.
A subscription adds:
- The advanced AI models: Whisper Large Turbo, NVIDIA’s Parakeet v3, Whisper Medium and Whisper Small.
- Word (.doc) exports of transcripts and notes.
- WebVTT (.vtt) subtitle exports that keep your per-subtitle styling.
- Subprojects to organise larger work.
- Project archives to back up a project or send it to a colleague.
Need tags for coding, advanced search, a literature library or citations? They are part of our sister research suite, Qualitativist.
Why do local transcriptions matter?
Most transcription services run on remote servers, which means uploading your audio and video files to someone else’s platform. That cloud-based approach usually costs you control over sensitive material, comes with opaque storage practices, and leaves you unsure how your content may be reused or retained.
Transcription App takes a fundamentally different approach. All transcriptions are performed locally on your device, using on-device AI models. No recordings are uploaded, no content is stored remotely, and nothing is shared with third parties. That makes the app a natural fit for professional, academic and creative work where confidentiality and data sovereignty are non-negotiable: protecting a journalist’s sources, honouring a research ethics agreement, or keeping an unreleased film under wraps.
Once your models are downloaded, transcription works without an internet connection, on a train, in the field or in an edit suite with no network. Your projects are saved as ordinary files in a folder you choose, so they are easy to back up, and subscribers can archive a whole project in one click whenever they need a copy.
An app, not a platform!
Transcription App is designed as a standalone application, not a platform. That choice is deliberate.
- No user accounts tied to your data
- No browser-based dashboards or cloud interfaces
- No uploads to third-party servers for processing
- No reuse of your content for training or analytics
Everything runs locally, on your machine. You install the app, work on your files, and keep full control over your material, without being locked into an ecosystem or service layer. Your licence is linked to the devices you verify, and you can unlink a device from the settings whenever you change machines.
Frequently asked questions
Does my audio ever leave my computer?
No. Transcription, speaker detection and speaker recognition all run on your device. Your recordings and transcripts are stored in a local folder and are never uploaded anywhere.
Do I need an internet connection?
Only occasionally. The app connects to check your licence and to download a model the first time you use it. Once a model is on your computer, you can transcribe completely offline.
Which languages are supported?
Whisper models transcribe 99 languages. You set the language for each file when you add it, and you can choose a default language in the settings so you do not have to pick it every time.
Which files can I transcribe?
Most common audio and video formats, including MKV. Video files are transcribed from their audio track, and very large files are converted automatically before transcription begins. You can also import existing transcripts and subtitles in SRT, VTT, TXT, DOC, HTML or JSON format.
How accurate is it?
Accuracy depends on the recording quality, the language and the model you choose. The demo above shows unedited results from four models on the same file, and the free mode lets you test complete recordings of your own with Whisper Base before you subscribe.
What does my computer need?
Transcription App runs on macOS, Windows and Linux. Macs with Apple silicon use their graphics processor for fast transcription. On any computer, smaller models such as Whisper Small and Whisper Base are the best choice when memory is limited, and the app falls back to a compatible setting automatically if a model cannot load.
Can I share a project with a colleague?
Yes. With a subscription, export the project as a single archive containing its recordings, transcripts, notes and subtitle presets, then send it however you prefer. Your colleague imports it into their own copy of the app. In the free mode, you can share exported transcripts and subtitles.
What is included in the free mode?
Everything you need to transcribe real work, free forever and with no time limit: complete recordings of any length with Whisper Base, unlimited projects, the queue, speaker segmentation and recognition, the transcript and subtitle editors, notes and search, exports to .txt, .srt and .json, transcript imports and full-storage backups. A subscription adds the more advanced models, subprojects, project archives, Word (.doc) exports and WebVTT (.vtt) exports. If a subscription ends, your work stays accessible.
Your recordings, your computer, your transcripts
Download Transcription App, transcribe your first recording in minutes, and keep every word under your control.