Skip to content
AI

Dictation

Free during Beta

Dictation turns speech into clean, formatted text anywhere you type. Trigger with a hotkey, speak naturally, and Raycast removes filler words, fixes punctuation, and pastes the result instantly. On iOS, the Raycast Keyboard brings the same experience to every text field.

The Dictation Pill floating on a wallpaper

When you first open Dictation, you'll get started in three steps:

  1. Grant Microphone access and Accessibility permissions so transcriptions paste directly into your focused app (without Accessibility access, they're copied to your clipboard)
  2. Pick your input device. We recommend the built-in or wired microphone for the lowest latency.
  3. Set your hotkey to trigger Dictation from anywhere without opening Raycast first

Want a single / fn key as your dictation hotkey? Open System Settings → Keyboard → Press / fn key to and set it to Do Nothing so the system doesn't intercept the press first.

Press your hotkey to start a session. The Dictation Pill appears above your current app, showing a live waveform and timer. Speak naturally as Raycast handles filler words, punctuation, capitalisation, and grammar automatically.

Press your hotkey again to accept or Esc to cancel. Your text pastes into the active app, or copies to your clipboard, depending on your settings.

Dictating a grocery list into a Raycast Note, with the Dictation Pill showing a live waveform at the bottom of the screen

Prefer hold-to-talk? Hold the hotkey down. Raycast switches to push-to-talk and when you release the hotkey, your session ends and recording finishes automatically.

Personalization is where Dictation starts to feel like yours. Fine-tune every transcription with custom instructions, app-aware context, and a vocabulary built around the words and names that matter to you. You can personalize your experience in Raycast Settings → Dictation -> Personalization.

Add global guidance to shape how your words are transcribed. Use it to enforce things like spelling preferences, grammar rules, tone, name capitalisation, and formatting conventions. These instructions sit above any per-style settings and apply to every transcription.

With App Context, Raycast reads the frontmost app, including its name, the field you're focused on, and any visible text nearby, then passes that to the transcription model to improve transcription accuracy.

App Context is used only for that transcription request and is never stored. Once the transcription is complete, the context data is discarded.

Add words, names, brand terms, or jargon that Raycast should always get right. Entries are passed to the transcription model to improve accuracy when it hears something close to them.

Styles shape your dictated text for different contexts. One spoken sentence becomes a tight Slack one-liner, a properly punctuated email, or any other format you define. You can control your transcription styling in Raycast Settings → Dictation → Styles.

Auto Styling picks the right style for the app or website you're dictating into. Email style in Mail, Messaging style in Slack, or a custom "Code Comments" style in your editor. When Auto Styling is off, Raycast still applies baseline cleanup but you don't get app-aware formatting on top.

Two styles come built in to Raycast. Email can format greetings, sign-offs, add paragraph breaks, and full punctuation. Messaging keeps transcriptions short, casual, and lightly punctuated for communication apps.

You can adjust which apps and websites trigger each built-in style. If you use an app or website for email that isn't in our list, add it to the Email style to get the same formatting.

Create your own style using the Create Style option in Raycast Settings, or use the Create Style command and give your style a Name & Icon, Prompt, and a list of Apps & Websites it should apply to.

The Create Style dialog with a custom style named 'steph style', a prompt with formatting rules, and a list of apps and websites it applies to

A style prompt tells the AI how to clean up your transcript to match your expected output. Describe the artifact (meeting notes, agenda, PR review), the voice, and spell out mechanics (punctuation, capitalization, numerals, emojis).

Use the Dictate to Note command to skip dictating to the active app entirely and save the transcription as a new Note. Useful for capturing ideas, voice memos, or meeting takeaways that you want to keep alongside your other notes without using another app or website.

Every transcription is saved to the Dictation History command so you can find and reuse it later. You can use the actions in the Action Panel to either Paste, Copy to Clipboard, Delete Transcription for the selected transcription, or Delete All Local Transcriptions to remove them all.

The Dictation History command showing a list of recent transcriptions grouped by Today and This Month, with details for the selected entry on the right

Use the Copy Last Transcription command to quickly copy the last transcribed text without needing to open the Dictation History command.

All Dictation settings live under Raycast Settings → Dictation, including your preferred input device, general behavior options, and usage statistics.

Raycast Settings open on the Dictation pane, showing usage statistics, microphone preferences, and general behaviour options

Over time, Dictation builds up a picture of your usage through statistics. You'll see your average WPM, estimated time saved, and total words dictated. Your stats are computed entirely on your device from your local transcription history. Statistics are not sent to Raycast as analytics nor telemetry.

By default, Raycast uses your System Default microphone from macOS System Settings. You can customize your microphone priority list to switch between input devices throughout the day.

Use to reorder microphones by priority, or press to exclude a microphone from the list. Your currently active microphone appears in green at the top.

  • Language: Select a specific dictation language, or leave on Auto to detect what you're speaking each time.
  • Output Action: Paste the transcript directly into your active app (Default), or copy to your clipboard.
  • Sound Effects: Play feedback when dictation starts or stops.
  • Mute While Recording: Mute system audio while a session is active.
  • Allow Enter Hotkey: Use to accept the transcription in addition to your Dictate hotkey.

Dictation lives inside the Raycast Keyboard and enables you to dictate from most text field in apps on your iPhone or iPad.

Dictation listening in the Raycast iOS app, with the Raycast AI Dynamic Island activity at the bottom of the screen

Install and enable the Raycast Keyboard before using Dictation. See Keyboard for setup instructions.

  1. In any app, tap a text field and switch to the Raycast Keyboard using the icon.
  2. Tap the button in selector on the Keyboard, then tap to begin dictating. At the beginning of a dictation session, you will be redirected to the Raycast app – this is necessary since iOS keyboards don't have direct microphone access.
  3. Speak naturally. Raycast handles filler words, punctuation, capitalisation, and grammar automatically.
  4. Stop the recording with or from the Live Activity. The cleaned-up text is inserted into your text field, or held for review, depending on your Insert Mode setting.

As of iOS 26.4, Apple removed the ability for keyboards to detect the host app's bundle ID, so after dictating you'll need to swipe back to your previous app manually. We hope this is resolved in a future iOS update.

After you start a dictation session, Raycast keeps a 5-minute session alive in the background so you can dictate multiple times without recording to the app each time. While a session is active, you'll see a Live Activity. Tap it to control or end the session. You can also end it from the main app by tapping the power button.

Don't want a background session? Toggle Disable Session in Raycast for iOS Settings > Dictation to end it immediately, or after 5 minutes, 10 minutes, 15 minutes, 30 minutes, or 1 hour.

By default, the dictation is inserted into your text field automatically. You can change the insertion mode in completing the following settings:

  1. Open the Raycast app.
  2. Tap your profile in the top right.
  3. Tap Keyboard then change Insert mode to one of the following; Auto-inset, Review first, or Review on AI command.

Post-Processing

iOS Dictation includes the same concept as Styles, and allows you to run your dictation through a post-processing flow on top of the standaring dictation cleanup to modify the output before inserted.

You can select the post-processing option under the Dictate button in the Raycast Keyboard, and choose from one of the following options:

  • Email: Format your output for email
  • Notes: Format your output for note-taking
  • Custom AI Commands: Your own prompts, ideal for personal writing styles, translation, or specific tone adjustments