Now live on iOS - Download

Download iOS

Voice Dictation for Xcode (2026)

Xcode 26 added a Swift mode to Voice Control, so on macOS Tahoe you can now dictate Swift and drive the IDE by voice. But it is Swift-only, Xcode-only, and built for hands-free control rather than fast everyday dictation. For code identifiers, the terminal, commit messages and talking to AI assistants, a system-wide layer like Aqua fits the way iOS work actually happens.

That is the short version. Below: what Xcode 26 genuinely ships today, where its built-in voice stops, and how to dictate across the whole iOS workflow instead of just the editor.

Does Xcode have built-in voice dictation? Yes, sort of, as of Xcode 26

This is the editor where the honest answer is closest to yes, so it is worth being precise. At WWDC 2025 Apple added a programming mode for Swift to Voice Control, its accessibility system, and shipped it in Xcode 26 on macOS Tahoe. With it you can speak Swift and have Voice Control understand the syntax: it handles spacing, operators and camelCasing, and it lets you navigate and edit around the IDE by voice. Apple documented it in the "What's new in Xcode 26" session. It is a real, meaningful accessibility advance, and if you code hands-free it is the biggest Apple has shipped in years.

Two things about it decide whether it is the right tool for you.

First, it is Voice Control, not free dictation. Voice Control is a command-and-control grammar built so people who cannot use a keyboard can operate the entire Mac by voice. Dictating a line of Swift through it means the system is parsing your speech as structured commands and code tokens, which is powerful for full hands-free operation and heavier than you want if you just wish to talk a sentence into a box. Developers trying it for casual dictation report the familiar friction of a general recognizer treating code as ordinary English until you lean on the command vocabulary.

Second, and more limiting for most people: it is Swift, inside Xcode. That is a big surface if your whole day is Swift in the editor. It is a small one the moment you step outside it, which in modern iOS work is constantly.

Where Xcode's built-in voice stops

Think about an actual afternoon shipping an iOS feature. You are in the Swift editor, yes. You are also in the integrated terminal running xcodebuild, xcrun simctl, pod install or a fastlane lane. You are writing a commit message, then a pull request description on GitHub in the browser. You are dropping a note to a teammate in Slack. And increasingly you are typing into an AI assistant, because Xcode 26 also shipped built-in Coding Intelligence with ChatGPT and support for other model providers, and plenty of iOS developers keep Claude Code or Cursor open alongside for the non-Swift parts of the project.

Voice Control's Swift mode helps with exactly one of those: Swift in the editor. It does nothing for the terminal, the commit box, the PR description, Slack, or the assistant prompt. Those are where a growing share of the words you produce now live, and they are prose, not Swift syntax, which is the input voice is fastest at in the first place. The built-in feature optimized the one surface that was already the most structured, and left the conversational surfaces on the keyboard.

That is the gap a system-wide layer fills. Not a replacement for Voice Control if you rely on hands-free control, but the tool for everything Voice Control's Swift mode does not reach.

What voice dictation for the whole iOS workflow has to do

Four requirements, given the built-in feature covers only Swift-in-Xcode:

  1. Apple-stack identifier accuracy. It has to write SwiftUI, @StateObject, NSAttributedString, URLSession, CoreData, viewDidLoad, xcodebuild and .xcodeproj as written, not as a phonetic guess. Aqua runs its own proprietary model, Avalon, trained on human-computer-interaction speech (prompts, code, commands, email) rather than audiobooks and podcasts. In a 9to5Mac side-by-side against built-in macOS Dictation on the same passage, Apple Dictation made 17 errors and Aqua made 1.
  2. Every surface, not just the editor. A system-wide layer inserts text at your cursor in the Swift editor, the integrated terminal, the commit box, the GitHub PR description in your browser, Slack, and an AI assistant prompt, without switching tools or modes.
  3. Context awareness. Aqua's Deep Context reads what is on screen, so when your editor is full of one project's type names, transcription leans toward those terms instead of their English near-homophones.
  4. Speed. Aqua inserts final text about 450ms after you release the key. Dictation that makes you wait is not worth the reach for the mic.

On independently checkable footing: Avalon placed #1 among proprietary models on the OpenASR leaderboard at its October 2025 debut. Aqua's own AISpeak benchmark, which we label self-reported, puts Avalon at 97.3% on AI and coding jargon. Lead with the checkable proof and treat the in-house number as directional. Aqua covers 49 languages and runs on Mac, Windows and iPhone.

How to set up voice dictation for Xcode

Because Aqua is an OS-level layer, there is nothing to install inside Xcode, no scheme to edit, and no setting to flip. You set it up once at the system level and it works in every Xcode surface and every app around it.

  1. Install Aqua for macOS and sign in.
  2. Grant accessibility / input permission when prompted so Aqua can insert text into other apps. That is System Settings, then Privacy & Security, then Accessibility. (This is the same permission Voice Control uses, and the two can coexist.)
  3. Pick your push-to-talk key. The default is Fn (hold to record, release to insert). Hold-to-talk suits an editor, where your bursts are short: a prompt here, a comment there.
  4. Click into any surface: a Swift file, the integrated terminal, the commit box in the Source Control navigator, a PR description in the browser, or your AI assistant.
  5. Hold Fn, speak, release. The text lands at the cursor. Read it, then run or send.

The same habit keeps working the moment you leave Xcode for a fastlane command or a Slack reply.

A real example

You just finished a view and want the assistant to wire it up. Click into the Coding Intelligence panel (or Claude Code in the terminal), hold Fn, and say:

"Add a URLSession call in WeatherService that fetches the forecast, decode it with Codable, and update the @Published property on the main actor."

Release, and the whole instruction lands with URLSession, Codable and @Published intact, ready to send. No reaching back to repair a mangled type name. That is the sentence-shaped, framework-dense input a general recognizer fumbles and a code-tuned one gets right.

Dictating code accurately: a few habits

Voice is for intent, not for spelling out every character:

  • Speak identifiers naturally. Say "url session" or "at published" and let a code-tuned model resolve the casing. For an oddly-cased symbol you use constantly, add it to your dictionary once instead of fighting it every time.
  • Let the assistant handle boilerplate. Describe the goal ("decode the response with Codable and update on the main actor") rather than narrating braces and semicolons. The model writes the Swift; you supply the direction.
  • Add your constants to the Custom Dictionary. Framework names, your own type and module names, a target scheme, a teammate's handle (up to 800 entries on Pro) so they land correctly every time.
  • Use Custom Instructions for standing preferences, for example "keep commit messages in conventional-commits format."

You will still type the occasional exact token, and that is fine. Most of what you produce in a day, PR descriptions, commit messages, comments and assistant prompts, is reasoning and instruction, which is what voice is fastest at. The same logic drove our write-ups on voice dictation for Claude Code and speech-to-text for Cursor.

A quick disambiguation: dictating in Xcode vs building dictation into your app

Two different searches collide on this query, so let us split them.

If you want to talk to Xcode and the tools around it, this post is for you: Voice Control's Swift mode for hands-free coding, plus a system-wide layer like Aqua for everything else.

If instead you are an iOS developer who wants to build speech-to-text into your own app, you want Apple's frameworks, not a dictation tool. That is the Speech framework, historically SFSpeechRecognizer, and in iOS 26 the newer SpeechAnalyzer and SpeechTranscriber APIs for on-device transcription. Different job entirely, and Apple's developer documentation is the right starting point. If you landed here for that, no dictation app is what you need.

The bottom line

Xcode 26 shipped the most credible built-in voice feature of any editor we have covered, and it is genuinely good at what it is for: hands-free Swift, inside Xcode, through Voice Control. It just covers one surface of a workflow that now spans the terminal, the commit box, the PR description, Slack and an AI assistant, and those conversational surfaces are exactly where dictation pays off most. A system-wide, code-tuned layer handles all of them, gets your Apple-stack identifiers right, and keeps working in every app. We have typed for 150 years. In a workflow you increasingly talk your way through, it is time to speak.

On price, Aqua Pro is $8/mo billed annually, with a free Starter tier (1,000 words) to try it and a 70% student discount on a .edu email. To build voice into your own tooling, the Avalon API is an OpenAI-compatible endpoint at $0.39 per hour of audio. See how Aqua compares to Wispr Flow, and the full feature tour is in the Aqua guide.

One honest caveat: Aqua is cloud-based, so it needs a network connection. On a fully offline machine, a local tool wins. For everyday connected work, the code accuracy and the everywhere-reach are worth the trade.

FAQ

Does Xcode have built-in voice dictation? Partly, as of Xcode 26 on macOS Tahoe. Apple added a Swift programming mode to Voice Control, its accessibility system, so you can dictate Swift and navigate the IDE by voice. It is Swift-only and works inside Xcode, and it is a command-and-control grammar rather than free-form dictation. For the terminal, commit messages, PR descriptions and AI prompts, you still need a system-wide layer.

How do I dictate code in Xcode? For hands-free Swift, enable Voice Control in System Settings and use its programming mode inside Xcode. For fast everyday dictation across the whole workflow, install a system-wide layer like Aqua on macOS, grant accessibility permission, then hold the Fn key, speak, and release. Text is inserted at your cursor in the Swift editor, the terminal, the commit box, and every app outside Xcode.

What is the most accurate speech-to-text for Xcode? Accuracy on Swift identifiers, framework names and CLI commands is what decides it. Aqua runs its own Avalon model, trained on prompts and code rather than audiobooks, and placed #1 among proprietary models on the OpenASR leaderboard at its October 2025 debut. In a 9to5Mac test it made 1 error where macOS Dictation made 17.

How much does Aqua cost for iOS developers? Aqua Pro is $8/mo billed annually, with a free Starter tier (1,000 words) and a 70% student discount with a .edu email. There is also an Avalon API at $0.39 per hour of audio for building voice into your own apps and developer tooling.