My regular cycle is to talk informally to the CLI agent and ask it to “say back to me what you understood”, and it almost always produces a nice clean and clear version. This simultaneously works as confirmation of its understanding and also as a sort of spec which likely helps keep the agent on track.

UPDATE - just tried handy with Parakeet v3, and it works really well too, so I'll use this instead of VoiceInk for a few days. I just also discovered that turning on the "debug" UI with Cmd-shift-D shows additional options like post processing and appending trailing space.

I do, however, wonder if there is a way all these TTS tools can get to the next level. The generated text should not be just a verbatim copy of what I just said, but depending on the context, it should elaborate. For example, if my cursor is actively inside an editor/IDE with some code, my coding-related verbal prompts should actually generate the right/desired code in that IDE.

Perhaps this is a bit of combining TTS with computer-use.

Superwhisper — Been using it a long time. It's paid with a lifetime subscription available. Tons of features. Language models are built right in without additional charge. Solo dev is epic; may defer upgrades to avoid occasional bugs/regressions (hey, it's complex software).

Trying each for a few minutes:

Hex — Feels the leanest (& cleanest) free options mentioned for Mac in this thread.

Fluid Voice — Offers a unique feature, a real-time view of your speech as you talk! Superwhisper has this, but only with an online model. (You can't see your entire transcript in Fluid, though. The recording window view is limited to about one sentence at a time--of course you do see everything when you complete your dictation.)

Handy — Pink and cute. I like the history window. As far as clipboard handling goes, I might note that the "don't modify clipboard" setting is more of a "restore clipboard" setting. Though it doesn't need as many permissions as Hex because it's willing to move clipboard items around a bit, if I'm not mistaken.

Note Hex seems to be upset about me installing all the others... lots of restarting in between installs all around. Each has something to offer.

---

Big shout out to Nvidia open-sourcing Parakeet--all of these apps are lightning fast.

Also I'm partial to being able to stream transcriptions to the cursor into any field, or at least view live like Fluid (or superwhisper online). I know it's complex b/c models transcribe the whole file for accuracy. (I'm OK with seeing a lower quality transcript realtime and waiting a second for the higher-quality version to paste at the end.)

P.S. The post processing that you are talking about, wouldn’t it be awesome.

Handy first release was June 2025, OpenWhispr a month later. Handy has ~11k GitHub stars, OpenWhispr has ~730.

The ui is well thought out, just the right amount of setting for my usage.

Incredible !

Btw, do you know what « discharging the model » does ? It’s set to never by default, tried to check if it has an impact on ram or cpu but it doesn’t seem to do anything.

How have your computing habits changed as a result of having this? When do you typically use this instead of typing on the keyboard?

Is there any way to execute commands directly on Linux?

Also a feature to edit or correct already typed text would be really great.