Voice turns into text in a fraction of a second.

Dictator puts the recognized text wherever your cursor is: an email, a chat, a document, a spreadsheet. A short phrase is ready in under a second. There is no limit on length.

• The default hotkey is Ctrl+Win: the first press starts recording, the second stops it. You can also hold the key like a walkie-talkie.
• Speech is recognized by a neural network. It runs on your own computer and understands 100 languages.
• It runs on NVIDIA, AMD and Intel graphics cards and on Apple M1-M5 chips. On a Mac it is as fast as on a graphics card.
• You can pause a recording. The cross cancels it, the check mark sends it.
• On the phone the Dictator keyboard works in any app: the mic sits above the letters, tap, speak, tap again, and the text is in the field. iPhone with iOS 15, Android with 7.0.

In translation mode the widget shows the target language flag. Click the flag to pick another language. You can paste the translation, what you said and the back-translation in any order.

• Offline translation into 45 languages. Free and without internet: dictionaries run right on your computer and phone.
• AI translation into 198 languages. The AI translates the meaning of the whole phrase, not word by word, so you are understood exactly. Quality is the same on every plan, only the volume differs.
• On the phone the 文 button on the keyboard records your speech and inserts the translation, the flag on the left picks the language. With text selected, the same button translates the selection.
• Back-translation. Dictator translates the result back into your language and shows it alongside. You can see at once whether the meaning came through, even if you do not know Chinese.

One and a half to two seconds from the start of speech to the spoken translation. The AI interpreter translates continuously and by voice, by the meaning of the phrase rather than word by word. For negotiations, travel, a doctor, talking to family abroad. Put the phone between you, each person gets their own half of the screen.

• Understands speech in 60-70 languages, translates by voice into 13-29 languages and with subtitles into 60. The exact number depends on the translation model you choose.
• The delay is the same on every plan, free minutes included. The translation can be voiced in a voice similar to yours.
• Modes: speakerphone, headphones, push to talk, subtitles only.
• 3 minutes free, 10, 30 and 60 minutes a month on paid plans, extra minute packs available.

Say or type a phrase: AI translation carries the meaning and the app reads it aloud. The Conversations tab shows who said what, and the conversation is kept in history.

• The back-translation button shows how the translation reads in your language: you check the meaning in a second.
• Languages are picked by flag, recent ones always on top.
• The translation is read aloud: a standard voice for free, neural voices on paid plans.
• One account on phone and computer: shared history.

On the phone the camera translates on the fly, and so do photos from the gallery. On the computer press Ctrl+Alt+W: select an area of the screen and get the translation. A game, an app, a picture where text cannot be copied.

• Live camera and photos from the gallery.
• A screen area on the computer is recognized right on it.
• You can ask AI about a translated photo: what this dish is, what the fine print says.

MP3, WAV, WebM, MP4, MOV, MKV. On your computer transcription is free, with speakers told apart.

• Rename speakers, add timestamps if you like.
• AI writes a short summary, removes filler words, rewrites the text in a business style or by your template. Every version is kept separately.
• On Windows, export to Word, PDF, SRT and VTT subtitles, TXT, Markdown and CSV; on a Mac, to TXT and Markdown.

Tables, headings and lists stay where they were. Compare the original and the translation line by line.

• Formats: PDF, DOCX, XLSX, PPTX, EPUB, HTML, MD, CSV, TXT.
• 3 documents a month free, 10 and 50 on paid plans, unlimited on Premium.

Add a client surname, a company name or a term to the dictionary. Dictator hints it to the model before recognition. With a template you say a short phrase and a long one is pasted: a signature, company details, a reply to a client.

• A dictionary word can be replaced with a different spelling right away.
• 10 words and 10 templates free, 30 on Basic, unlimited on Pro and Premium.

Voice a translation, an article, your own text. The sentence is highlighted as it is read.

• Free and unlimited: the standard and the system voice.
• On the computer, offline: a light voice with almost instant first sound, and a high-quality voice on the graphics card.
• Studio voices: cloud neural networks with natural intonation.
• A clone of your voice on Premium.

Set up a free shared server: it is simply your computer with Dictator, no separate server needed. Invite people by email, code or link. Phones and weak laptops get the recognition quality of a powerful graphics card.

Everyone recognizes speech fast, accurately, for free and privately.

• 2 people free, 4, 10 and 25 on paid plans.
• Recordings never leave your network: lawyers, doctors and sales teams value that.