← All guides

JetBrains and IntelliJ voice input on Mac

JetBrains Air has local voice input. Classic JetBrains IDEs are a different path.

JetBrains Air now documents a microphone in its Chat task input backed by a downloaded local Whisper model, with no audio sent to the cloud. The current classic IntelliJ IDEA editor and AI Assistant documentation reviewed here describes text, chat, context, and commit fields but does not identify that same native microphone. One guide can cover the shared decision without transferring Air’s privacy claim to IntelliJ, PyCharm, or WebStorm.

Source-reviewed

1. Separate JetBrains Air from classic IntelliJ-based IDEs

JetBrains Air is an agentic development environment with its own Chat task input. Its current documentation explicitly includes optional voice input and a dedicated Voice Input setting.

IntelliJ IDEA, PyCharm, WebStorm, and other classic JetBrains IDEs have editor, commit, and AI Assistant text surfaces. The current classic IDE documentation reviewed here does not identify Air’s microphone as a shared platform feature. Do not copy Air setup or local-audio claims into every JetBrains product.

2. Use Air’s native microphone for an Air task prompt

Open Air Settings, search for Voice Input, and download a local Whisper model. JetBrains lets you optionally choose the model size and recognition language. When the status is Ready, return to Chat and select the microphone in the task input.

JetBrains states that Air voice recognition uses the local Whisper model and no audio is sent to the cloud. That narrow statement covers the speech-recognition stage. After transcription, the prompt, selected context, agent, model, environment, permissions, and tool actions follow the Air task you configure.

  • Download the Air voice model before the first task.
  • Speak into the visible Chat task input.
  • Review the task, context, environment, agent, model, and permission mode.

3. Use focused-field Dictation in a classic JetBrains IDE

In IntelliJ IDEA, PyCharm, WebStorm, or another classic IDE, place the insertion point in the exact editor, documentation, comment, commit-message, or AI Assistant prompt field. Start the macOS Dictation shortcut, speak, stop, and confirm the transcript appeared in the intended surface.

JetBrains documents Escape for moving focus back to the active IntelliJ editor and distinct shortcuts for tool windows. Focus matters because a transcript intended for a comment can otherwise land in AI Chat, Search Everywhere, the terminal, or another custom editor.

4. Treat code, commands, and identifiers as high-risk text

Dictation is useful for prose in documentation, comments, commit messages, and agent requirements. It is not deterministic source-code entry. Check case-sensitive symbols, file paths, package names, generics, operators, punctuation, quotes, indentation, numbers, negations, and security-sensitive wording.

For exact code or shell commands, prefer typing or pasting from a trusted reviewed source. Do not press Return in a terminal or accept an AI-generated change merely because the transcript sounds plausible.

5. Keep AI Assistant context and agent actions separate from input

JetBrains AI Assistant supports chat, code-aware context, files, folders, symbols, images, and agent workflows across supported IDEs. Dictation can create text in an accepted prompt field; it does not choose the model, attach the right context, approve a tool call, apply an edit, or run a command.

In Air, JetBrains separately asks you to choose a task run environment, agent and model, permission mode, and context before running a task. Inspect those controls after voice recognition and before submission.

6. Review commit messages and troubleshoot focus separately

IntelliJ IDEA provides an explicit commit-message field and Commit action. Dictate the rationale if useful, then verify issue keys, component names, versions, breaking-change language, and the selected changes before committing or pushing.

If Dictation fails everywhere, test TextEdit and follow Apple’s checks for Keyboard settings, shortcut, language, microphone, Voice Control, and connectivity. If only one JetBrains surface fails, confirm it is editable, move focus deliberately, and use the clipboard fallback for a custom editor that rejects system insertion.

7. Add IraVoice only for the cross-app or structured brief workflow

Use JetBrains Air’s microphone first when the destination is already an Air task prompt. Use macOS Dictation first for occasional prose in a supported classic IDE field.

IraVoice 0.7.1 is the separate option for macOS 26+ Apple-silicon Macs when one held-key workflow should produce editable text across supported apps, or when a longer spoken engineering request should become reviewable Markdown before you choose an IDE or agent. IraVoice does not read a JetBrains project, choose context, control AI Assistant or Air, apply edits, run code, press Commit, or push changes.

Decision guide

Which JetBrains voice-input path fits?

Choose by product surface, text risk, and whether an agent can act afterward.

If you needChooseWhy
Speak an agent task in JetBrains AirAir Voice InputIt is the native task-input microphone with documented local Whisper recognition.
Dictate prose in IntelliJ, PyCharm, or WebStormmacOS DictationIt can provide text to a supported focused editor, comment, commit, or chat field.
Use one local voice workflow across several Mac appsIraVoice DictateIt is cross-app rather than tied to one Air task input.
Create a reusable engineering brief before choosing an IDEIraVoice Voice-to-SpecIt produces reviewable Markdown before any agent receives the requirement.
Enter exact code, commands, or identifiersType or paste, then reviewSpeech recognition should not be treated as deterministic syntax entry.

Primary sources

Use the current documentation.

This guide was reviewed July 26, 2026. Product versions, settings, policies, and availability can change.

  1. JetBrains Air: Define tasks and configure Voice input

    Current Air Chat task workflow, microphone, local Whisper model, no-cloud-audio statement, context, environment, agent, model, permissions, and run boundary.

  2. JetBrains: IntelliJ IDEA editor basics

    Current editor purpose, focus changes, terminal-to-editor navigation, tool windows, clipboard, and inspection workflow.

  3. JetBrains: About AI Assistant

    Current supported IDEs and AI Assistant chat, editor, model, code, and agent capabilities.

  4. JetBrains: AI Chat

    Current text prompt, context attachment, files, folders, symbols, images, response, and agent workflows.

  5. JetBrains: Commit and push changes

    Current commit-message field, change review, Commit action, and push workflow.

  6. Apple: Dictate messages and documents on Mac

    The separate system Dictation shortcut, focused-field workflow, stop controls, and configuration-specific processing check.

  7. Apple: If Dictation on Mac does not work as expected

    Current Keyboard, shortcut, language, microphone, Voice Control, connection, and app-field troubleshooting.

Questions

What readers usually ask next.

Does IntelliJ IDEA have voice input?

The current classic IntelliJ IDEA editor and AI Assistant documentation reviewed here does not identify the native JetBrains Air microphone as a classic IDE feature. On a Mac, use system Dictation in a supported focused IntelliJ field, then review the text.

Does JetBrains Air have voice input?

Yes. JetBrains documents a microphone in Air’s Chat task input after you download and configure a local Whisper model under Air Settings → Voice Input.

Is JetBrains Air voice input local?

JetBrains says Air voice recognition uses a local Whisper model and no audio is sent to the cloud. The resulting task, context, model, agent, and tool work still follow the Air configuration you run.

Can I dictate code into IntelliJ or PyCharm?

A supported focused field may accept macOS Dictation, but speech recognition is not deterministic code entry. Review symbols, paths, punctuation, case, indentation, commands, and every consequential action.

Will dictation send or execute a JetBrains AI task?

Speech-to-text should create reviewable prompt text only. Sending, choosing context, selecting an agent or model, approving permissions, applying edits, running commands, and committing are separate JetBrains actions.

Can IraVoice control IntelliJ, PyCharm, or JetBrains Air?

No. IraVoice does not read projects, select context, control JetBrains agents, apply code, run tools, commit, or push. It can create local editable text or reviewable Markdown for you to place in the intended field.

macOS 26+ · Apple-silicon

Try the local Mac workflow for yourself.

Dictate ordinary text or turn a rough technical thought into reviewable Markdown during the 14-day full trial.