MachinesFluent v1.1.0 improves the daily dictation loop.
The update adds two cloud speech engines, makes Hands Free work with Smart Dictation, introduces vocabulary snippets, and cleans up several preview and settings details that affect normal use.
What changed in v1.1.0
| Area | What you can do now | Why it matters |
|---|---|---|
| Speech engines | Use Soniox and ElevenLabs as real-time cloud speech options. | More engine choice makes it easier to find the best fit for your voice, language, and microphone. |
| Local dictation | Faster Whisper and Parakeet handle pauses and quiet endings more reliably. | Final words are less likely to disappear or arrive awkwardly late. |
| Hands Free Smart Dictation | Tap once to record, then tap again to send the result through the selected AI prompt. | Longer spoken thoughts can be cleaned or structured before they land. |
| Vocabulary snippets | Expand a short spoken trigger into a longer saved phrase. | Repeated phrases, names, signatures, and templates are easier to dictate. |
| Preview and settings | Transcription Preview, Speech settings, AI Preview, provider menus, and onboarding copy are clearer. | The app gives better feedback while you work. |
More speech engines means more room to choose
Speech recognition is not one-size-fits-all. Accent, language, microphone, room noise, and the kind of text you dictate all change the result.
v1.1.0 adds Soniox and ElevenLabs as cloud speech options. It also improves local behavior in Faster Whisper and Parakeet, especially around pauses, weak endings, and release timing.
The point is practical: if one engine does not feel right for a job, you have more ways to test another.
Hands Free now works with Smart Dictation
Hands Free used to be mostly about live dictation. In v1.1.0, it also works with Smart Dictation.
When Smart Dictation is on, the first tap starts recording and the next tap finishes the capture. MachinesFluent then sends the completed transcript through the selected prompt before inserting the output.
That is useful for longer thoughts. You can speak the rough version first, then let the prompt turn it into an email reply, a cleaner note, a structured message, or whatever format you selected.
Vocabulary snippets remove repeated typing
Vocabulary snippets are for phrases you say often.
You set a short spoken trigger, and MachinesFluent expands it into a longer saved phrase after transcription. It does not require an AI call. It also does not depend on provider-side vocabulary hints.
Use it for product names, email signatures, standard replies, repeated boilerplate, or terms that speech engines keep hearing incorrectly.
Preview and settings are easier to trust
v1.1.0 also cleans up several parts of the app that are not flashy but matter during a work session:
- AI Preview handles normal dollar amounts without treating them like math.
- Transcription Preview has clearer modes for Press & Hold, Hands Free, or both.
- Speech settings are easier to scan.
- AI provider menus show provider icons.
- Lite builds make the included local Whisper models clearer.
The update makes the active mode easier to verify: preview mode, provider icons, and the local/cloud speech grouping now show what the app is doing before you dictate.
Should you update?
Yes, especially if dictation is part of your daily work.
Start by testing the new speech engines, then try Hands Free with a Smart Dictation prompt. If you repeat the same phrases often, add one vocabulary snippet and see whether it removes friction from your normal writing.
