Typing is the bottleneck
Commit messages, PR reviews, standup notes, and Slack replies eat the day one keystroke at a time — while the thought was already finished.
For developers, programmers & engineers
Dictation software
Dictation software
for developers
WhisperJot is a voice-dictation app for macOS, Windows, and Linux that types into whatever app has focus — VS Code, Cursor, a terminal, Slack. It is private by default: transcription runs on-device with the local engines and works fully offline, with an opt-in Jot Cloud engine on Pro if you prefer zero local model footprint. Local is $5/month; Pro is $99/year.
Why voice helps
The work is thinking, not typing.
Source code is confidential
Most dictation tools stream your audio to someone else's servers. What you say over proprietary code shouldn't leave the machine by default.
Linux is always the afterthought
Popular dictation apps ship Mac and mobile builds and skip the desktop where much of the actual engineering happens.
Example workflow
A workday on voice
One possible day with WhisperJot — adapt it to the tools and routines you already use.
- Morning: hold the hotkey and dictate standup notes straight into Slack — filler words stripped, punctuation handled.
- Reviewing a PR: talk through feedback into the GitHub comment box; 'new paragraph' and custom vocabulary keep function names exact.
- Deep work: dictate docstrings and README sections into VS Code with live partials, hands never leaving the keyboard for long-form prose.
- On the train, offline: the local engines keep transcribing — no connection, no degraded mode.
Built for the flow
Dictation that fits the way you work.
One hotkey in any editor
Press ⌥Space (Ctrl+Alt+Space on Windows/Linux) or hold fn, talk, and the text lands in VS Code, Cursor, Xcode, vim in a terminal — any focused app, with per-app overrides for how text is injected.
Custom vocabulary for jargon
Teach it your stack: Kubernetes, useEffect, psycopg2. Custom vocabulary correction keeps identifiers and product names intact.
Local by default, offline on a plane
Jot Local and Jot Local Pro run on-device — dictation keeps working with no connection. Jot Cloud is strictly opt-in, runs only while selected, and audio is never stored on our servers.
Clean text, not filler
Filler-word removal, automatic punctuation, and voice commands like 'new paragraph'. Optional LLM polish runs locally on 127.0.0.1.
History stays on disk
Every utterance lands in a local SQLite history you can search — grep-able, yours, on your machine.
An actual Linux build
Debian/Ubuntu builds for x64 and arm64 (beta), alongside stable macOS 14+ (Apple Silicon) and Windows 10+ x64 — all on either plan.
Questions
Straight answers.
Check the full system requirements or compare WhisperJot with other dictation apps. See the FAQ and the glossary.
Is dictation software safe to use with proprietary source code?
WhisperJot is private by default: with the local engines (the default), audio is transcribed on your machine and nothing is uploaded. The optional Jot Cloud engine is opt-in behind an explicit consent prompt, is used only while selected, and audio is never stored on our servers.
Can I dictate into VS Code, Cursor, or a terminal?
Yes. WhisperJot types into whatever app has focus, so it works in VS Code, Cursor, Xcode, JetBrains IDEs, and terminal editors, with per-app overrides for how text is injected. No plugin or extension is required.
Does WhisperJot handle technical jargon and variable names?
You can add custom vocabulary so terms like PostgreSQL, useEffect, or your internal product names come out spelled correctly. Filler-word removal and automatic punctuation keep the output paste-ready.
Is there dictation software that runs on Linux?
WhisperJot ships Linux builds for Debian/Ubuntu on x64 and arm64 (currently in beta), alongside stable macOS 14+ (Apple Silicon) and Windows 10+ x64 builds. Either plan covers all platforms — Local at $5/month, Pro at $99/year.
Does it work offline?
Yes — the local engines (the default) run entirely on-device, so dictation works with no internet connection. Only the opt-in Jot Cloud engine needs a connection.
Choose a setup and try a workflow.
Dictate actionable Linear tickets
Create a Linear issue draft by voice with scope, acceptance criteria and evidence. Use a team template, then review properties and submit through Linear.
Read the guide →Do you need custom dictation vocabulary?
Decide whether custom vocabulary will help your dictation. Compare name correction, text shortcuts and ordinary editing, with an example and practical limits.
Read the guide →Choose local text cleanup for dictation
Compare raw dictation, rule-based cleanup and optional local LLM polish. See setup requirements, an illustrative edit and checks for preserving meaning.
Read the guide →Move from cloud to local dictation
Move a dictation workflow to local engines with model setup, vocabulary checks and an offline test. Separate speech processing from sync and destination apps.
Read the guide →Check names and numbers in dictated text
Reduce correction work for dictated names, dates and reference numbers. Use a small vocabulary list and a source-checking routine before sharing drafts.
Read the guide →Give your hands a break.
One hotkey, any focused app, and private local transcription by default.