Help & Support
How can we help?
Email us at [email protected]. We read every message.
Getting started
Three things. Speak instead of typing, anywhere on your computer. Record and transcribe meetings, with speakers labeled. Turn audio and video files into text. Everything runs on your own machine.
Dictate is a Windows app. Download the installer from our website, run it, and sign in or create an account with your email. Any reasonably modern Windows computer works. A dedicated graphics card makes transcription faster but is not required.
When we release a new version, the app lets you know and you can install it from there. Your account, recordings, and settings stay as they are.
Dictation
Any app. Press the Dictate hotkey wherever you are typing, in an email, a document, or your browser, and your words land right there.
Yes. Live dictation shows the text as you speak, so you can watch it build and catch mistakes on the spot.
One word is one word in your finished transcript, whether it came from dictation, a meeting, or a file. Your 20,000 words reset every month.
Meetings
Yes. Dictate records the meeting, transcribes it live with captions, and saves it so you can search and manage everything afterward.
Dictate detects and labels speakers in a meeting, and you can profile a voice so the same person is recognized in later meetings. Speaker identification is in beta and still improving.
One meeting is one recording, from the moment you start it to the moment you stop it. Each counts toward your 10 monthly meetings, and the count resets every month.
Files
Almost any audio or video file. Add it to Dictate and the transcript is prepared on your machine, the same way as a live recording.
Yes. Transcripts export as SRT, VTT, or plain text. Meeting reports export as Word or PDF.
Yes. Files work in the same languages as dictation and meetings, including Indian languages.
AI features
You bring your own AI key for the provider you prefer. Requests go straight from your machine to that provider. Your key and that traffic never pass through us.
Auto-polish tidies raw dictation, things like punctuation, capitalization, and sentence breaks. One click gives you a meeting summary. You can also generate custom reports and export them.
Transcription works fully offline. AI features either run on your own machine or talk to your chosen cloud provider, and the cloud path needs internet.
No. Voice cloning is not part of Dictate today. If it ever ships, it will be built consent-first: cloning a voice will require the owner's explicit agreement.
Set up OpenAI is a short guide with every click: you create a key on their platform, add a little prepaid credit, and paste the key into Dictate. About five minutes.
One of the cheapest options. Set up DeepSeek walks you through creating a key and topping up a few dollars. About five minutes.
Anthropic makes Claude. Set up Anthropic shows the exact clicks, and new accounts include a little starter credit. About five minutes.
Ollama runs an assistant on your own computer, or on their cloud with no download at all. Set up Ollama covers installing it, adding a model, and the cloud trick. About ten minutes.
One key reaches models from many companies, including free ones you can start on with no card. Set up OpenRouter takes about five minutes.
Languages and speed
You can dictate and transcribe in your preferred language, with broad coverage of Indian and international languages. In meetings, mixed-language conversation is handled naturally.
When AI features run on your own machine, languages written in scripts like Hindi, Marathi, Bengali, Tamil, Telugu, Kannada, Malayalam, Gujarati, Punjabi, Odia, Nepali, or Thai can take two to four times longer than English. This is a property of how these writing systems are processed today, not a fault in your machine. Using a cloud AI provider avoids the slowdown completely, and on-device processing for these languages will keep improving.
Transcription runs on your computer, so speed follows your hardware. A dedicated graphics card makes it faster but is not required.
Plans and billing
The Free plan costs nothing: 20,000 words and 10 meetings, resetting every month. Ultimate is unlimited, at ₹299 per month in India or $7.99 elsewhere, ₹699/$18.99 per quarter, or ₹2,299/$59.99 per year. See the pricing page. Paddle, our Merchant of Record, handles payment, tax, and receipts, so your receipt always comes from Paddle.
Your word and meeting counts reset every month, so you can continue after the reset. If you need more before then, Ultimate lifts both limits. Your recordings and transcripts stay on your machine either way.
You can buy Ultimate any time, from the app or the pricing page. To change your plan or billing cycle, cancel and choose again, or write to us and we will help. Canceling stops the next renewal only, and you keep access until the end of the period you already paid for.
No. When you pay, you get the full period you paid for. That is exactly why the Free plan exists: test everything before you buy. Details are on our refund policy page.
Yes. One account can be signed in on up to three devices at the same time. Sign in with the same account on each. Recordings stay on the machine that made them.
Privacy and data
No. Dictation, meetings, and file transcription all run on your own machine. Your recordings and transcripts stay with you.
Your email, name, and avatar for your account, plus anonymous usage analytics hosted in the European Union, which you can switch off any time. We never sell your data.
Your recordings and transcripts are stored on your machine, encrypted per account. They are not kept on our servers, so signing in on a second computer does not copy them over. For a portable copy, export the transcripts and reports you care about.
Yes, from inside the Dictate app, in settings. Deleting erases your profile and synced content from our servers. Basic billing records are kept only as long as the law requires. Details are in our privacy policy.
Troubleshooting
First check that Windows itself can hear your microphone and that the right microphone is selected on your computer. Restarting the app clears most microphone problems. If it keeps happening, write to us.
Transcription follows audio quality. Get the microphone closer to whoever is speaking, cut background noise where you can, and check that your language is set correctly. Speaker labels are still in beta.
Write to [email protected]. Include what you were doing and what went wrong. We read every message, and the same address reaches us for abuse reports.
Still stuck?
Tell us what you were doing and what went wrong. One email, read by a person.
Email [email protected]