Voice input in Gemini: how dictation works

Potok transcribes Russian speech; this site has been translated into English for convenience.

Gemini's apps and website have dictation: the microphone icon in the prompt field turns speech into text at the cursor, and the recognition language depends on the chosen interface language. This is not the same as Gemini Live, a separate voice mode for talking with the assistant. You can also dictate text into any field through Potok (potok.now), about 900 characters per minute by voice.

Voice input in Gemini comes in two kinds: short dictation of a prompt with the microphone icon, and Gemini Live, a separate conversational mode. Let us look at how dictation works according to Google's help and what to do if you need to dictate a long text and not talk with the assistant.

How dictation works in Gemini

Gemini has built-in dictation: in the prompt field, on the web, in apps for Windows and Mac and in mobile apps for Android and iOS, there is a microphone icon. Pressing it turns speech into text right in the field where the cursor is, and the text can be corrected before the prompt is sent.

The recognition language depends on which language is chosen in the Gemini interface, and some voice features are available not in all languages, as stated in Google's help. Separate from this dictation there is Gemini Live: a voice mode for a natural conversation with the assistant, in which you can interrupt it on the fly. It is a different feature and not line-by-line dictation of text into the prompt field.

More about dictation is in Google's help: Use your voice with Gemini.

How to dictate text in Gemini through Potok

1. Open Gemini, in a browser, in a computer app or on a phone, and put the cursor in the prompt field or in any other text field.
2. Hold Potok's hotkey (Ctrl+Space by default) and, without releasing it, say the text.
3. Release the key: the recording goes to Potok's servers, the text is recognized and appears in the same field.
4. Check the text and send the prompt.

Punctuation and capital letters are placed automatically, and you do not need to dictate "zapyataya" (comma) or "tochka" (period) separately. Recognition is in the cloud, on its own model trained on Russian speech, including terms and mixed Russian text with English words, which is convenient for prompts in two languages. Potok states about 900 characters per minute against about 200 with ordinary typing. Potok's personal dictionary on the Pro plan remembers terms and names that often repeat in prompts. This way of dictating text works in other similar apps too, for example in ChatGPT or in Claude. If the default hotkey is inconvenient, it can be changed in the settings to a ready shortcut, your own key or a side mouse button.

When Gemini's built-in dictation is enough

If you need to quickly dictate a short prompt right into Gemini, the microphone icon in the prompt field is enough for this, and you do not need to install a separate app.

Potok is a cloud service, with recognition on servers in Russia; it does not work without the internet. It has no voice commands and no editing of text by voice: it only inserts the dictated text into the field you chose. It has no offline mode and no phone app: it is a desktop program for Windows, macOS and Linux.

Frequently asked questions

Built-in dictation and Potok

Primary sources: gemini.google.com · support.google.com · support.google.com · support.google.com