MachinesFluent v1.1.1 is mainly about control.
If you dictate into browser apps, switch between Hands Free and Press & Hold, or use History to recover past work, this update is worth installing. The big change is simple: Smart Dictation can now treat different websites as different workspaces, even when they all run inside the same browser.
What changed in v1.1.1
| Area | What you can do now | Why it matters |
|---|---|---|
| Browser apps | Assign Smart Dictation prompts by website or domain, not only by browser app. | Gmail, docs, dashboards, and admin tools can each get the prompt that fits the work. |
| Installed web apps | Route browser-installed PWAs separately from the main browser. | A web app pinned to the desktop can behave like its own workspace. |
| Recording controls | Switch between Hands Free and Press & Hold from the Speech Engines popup. | You can change recording style without opening Settings. |
| History retry | Retry an AI result from History with the saved request or current settings. | A useful dictation or AI pass is easier to recover instead of rebuilding it from scratch. |
| History retention | Keep History forever, for 24 hours, for 1 week, for 1 month, or never save it. | Saved transcripts and optional audio can match your comfort level. |
| Audio while recording | Optionally lower other app audio while the microphone is recording. | Background audio is less likely to fight your voice. |
Browser apps can use the right prompt
A lot of Windows work now happens inside the browser. That does not mean every browser tab is the same kind of work.
Writing in Gmail, updating a support dashboard, filling a web form, and drafting in a document editor all ask for different output. v1.1.1 lets Smart Dictation route prompts by website or domain, so the app can respond to the place where the work is happening instead of only seeing "Chrome" or "Edge."
That means Gmail, internal notes, admin fields, and a browser-installed web app can each use the prompt that matches the job.
Recording mode is easier to change
Hands Free and Press & Hold are for different moments.
Press & Hold is better when you want to drop one sentence into a field and move on. Hands Free is better when the thought is longer and you do not want to keep a key pressed. In v1.1.1, that choice is available from the Speech Engines popup, closer to the place where dictation starts.
There is also an optional volume-ducking setting for microphone recording. If another app is playing audio, MachinesFluent can lower it while you speak and restore it afterward.
History is more useful when an AI pass is almost right
v1.1.1 adds AI retry from History. You can retry with the saved request details when they are available, or rerun the entry with your current settings. The result goes to the preview or clipboard instead of pretending to be a brand-new dictation.
That matters when the first AI result was close but not quite there. Instead of copying old text, finding the right prompt, and rebuilding the request by hand, you can start from the saved entry.
Retention is now an explicit choice
v1.1.1 adds clearer retention choices: forever, 24 hours, 1 week, 1 month, or never. Optional audio sidecars follow the same cleanup rules when they apply.
If you use History as a recovery tool, keep it longer. If you only want short-term safety, use a shorter window. If you do not want persistent History, turn it off.
Smaller fixes worth knowing
The rest of the update is mostly friction removal:
- Smart Dictation prompt cards stay visible but locked when Smart Dictation is off, so you can still review the setup.
- Popup menus behave more predictably when opened repeatedly.
- Speech settings separate local and cloud models more clearly.
- AI provider errors and blocked preview states are easier to understand.
- Vocabulary snippets handle punctuation around triggers more cleanly.
These are not the headline, but they support the same goal: fewer small interruptions while using voice as a normal input method.
Should you update?
Yes, especially if you use MachinesFluent in browser-based work or rely on History to recover text and AI output.
After updating, test three things: assign a Smart Dictation prompt to a website, switch recording modes from the Speech Engines popup, and retry one History entry. That is where v1.1.1 should be most visible.
