A live voice keyboard for iPhone and Mac
The useful difference is not just that you can speak instead of type. It is that you can see, understand, and edit the words while they are still arriving.
Most voice dictation follows a batch loop: speak a thought, wait for processing, read a completed transcript, then discover what needs correction. That can work for a short command. It becomes expensive when the text is a customer reply, a product note, an email, a prompt, or a bilingual message with names and technical terms.
A live voice keyboard takes a different route. It lets text appear in the current writing surface as you speak. The words remain editable, context can repair common errors while the thought is forming, and the final step happens after you have already seen the important parts.
Why streaming changes the dictation loop
| Batch dictation | Live voice keyboard |
|---|---|
| Speak, then wait for a complete result. | See useful text while speaking. |
| Read the whole transcript after processing. | Notice important words and context as they appear. |
| Correct after the model has finished. | Let a final pass settle a result built on an already-visible draft. |
The win is not a promise of perfect recognition. It is a better place to put your attention. Instead of waiting for a paragraph and then rereading it from the beginning, you can understand the draft in motion and keep control of the text field you are already using.
What Echo adds to live dictation
- Stream first: editable text appears while you speak.
- Repair live: context helps resolve common homophones, boundaries, omissions, and punctuation while the draft forms.
- Finalize clean: a short pause settles the result instead of starting another long batch wait.
- Learn vocabulary: save names, product terms, and specialist language in a personal dictionary so future sessions fit your context better.
- Polish when needed: use optional formatting, rewriting, or translation after the usable transcript is already there.
Where a keyboard-level approach matters
Voice input is most valuable where the writing already happens: messages, email, notes, chat, prompts, and work tools. On iPhone, Echo is a voice keyboard for compatible text fields. On Mac, it is a quick input path for the same everyday writing loop. That is a different job from recording a meeting or creating a separate voice memo that needs to be moved elsewhere later.
Some system-controlled secure fields limit what third-party keyboards can do. For normal writing fields, the goal remains simple: speak naturally, watch the text become usable, and edit without changing tools.
Who benefits most
- People who write many short replies and do not want to type every sentence.
- Founders, engineers, PMs, sales teams, and creators turning rough thoughts into daily written output.
- Chinese-English bilingual users who switch languages inside the same thought.
- Anyone who likes the speed of dictation but dislikes cleaning up a finished transcript.
Echo's product thesis is concise: stream first, repair live, finalize clean, and learn your vocabulary.