A keyboard does not ask you to think about infrastructure. You press a key and a letter appears where you are working.
Voice input should feel just as dependable. Speaking should be another way to put a thought into an email, a prompt, a note, or a document—not a separate destination you have to visit.
The short answer
Voice input should behave like part of the computer. It should work where the cursor works, give you a clear local option, and make cloud processing a visible choice.
The goal is not to replace the keyboard. It is to make speaking a normal input method when speaking is the faster way to express the thought.
Input, not a destination
Most people do not wake up wanting a transcript. They want to reply to an email, explain a bug, write a note, or give an AI agent enough context to do useful work.
That is why voice becomes more useful when it works across the system. If you must record in one app, wait, copy the text, switch windows, and paste it somewhere else, voice has created another workflow instead of removing one.
Cue is built around a simpler idea: dictate where you are already typing. The thought starts in your voice and lands in the tool where the work continues.
Local and cloud should be a choice
Local-first does not mean pretending the cloud has no value. Cloud models can be useful when you want a different model, language coverage, or connected AI work.
But an everyday input method should not become useless only because a connection disappears. Cue lets people choose between on-device and cloud transcription and makes the active boundary visible.
The important part is not declaring one side good and the other bad. It is knowing which one you are using.
The words are only the beginning
Once a thought becomes text, the work continues.
Sometimes the text belongs in the app under the cursor. Sometimes it should become a Markdown note. Sometimes it becomes context for Claude, Cursor, Codex, or another tool through permissioned MCP access.
This is where voice stops being a dictation feature and starts becoming part of a workspace. Capture once. Keep using the thought.
When typing still wins
Voice is not better for everything.
Typing is usually better for exact commands, short corrections, private rooms, tables, and text where every character matters. A keyboard is quiet, precise, and easy to scan while you compose.
The useful product does not force a choice. It makes switching between voice and typing feel ordinary.
Frequently asked questions
Is voice input supposed to replace a keyboard?
No. It should sit beside the keyboard. Speak when the thought is easier to say; type when precision or privacy matters more.
Does Cue work only inside its own editor?
No. Cue's dictation is designed to work in Mac apps where you can type. Cue also includes Notes for work you want to keep in your own workspace.
Is local transcription always better?
No. The best option depends on the model, language, microphone, environment, and job. The useful difference is that the user can choose.
Sources
- Cue: an AI workspace built around your voice — accessed August 20, 2026.
- Cue Dictation — accessed August 20, 2026.
- Cue Notes — accessed August 20, 2026.
- Cue MCP — accessed August 20, 2026.
Try it in real work
Try Cue for Mac when you want to speak a thought directly into the app where you are already working. If you are new to the category, start with Free voice-to-text for Mac.