Launched this week
Loqui
Turns your voice into ready-to-send text in any Mac app
88 followers
Turns your voice into ready-to-send text in any Mac app
88 followers
Loqui turns your voice into clean, ready-to-send text in any app on your Mac. Hold a shortcut, speak, and the text lands at your cursor in the tone you pick. It also transcribes meetings and writes the minutes. 90+ languages. Free plan, no card needed.
Free Options
Launch Team / Built With
Framer AI AgentsDesign and publish professional sites with AI
Promoted
Hi Product Hunt! I'm Sergio, the maker of Loqui.
A while ago I came across apps that turn speech into polished text and loved the idea. I thought I could build one in that spirit, with the features I'd actually want. A few friends joined in to test every build, and after months of work (and a fair amount of fear of hitting publish), here it is.
What Loqui does:
Dictate in any app: hold a shortcut, speak, and clean text lands at your cursor
Tones: the same sentence comes out casual for a chat or polished for an email
Meetings: transcription, who said what, and ready-made minutes
90+ languages, with automatic detection
Your audio is discarded right after transcription, and your history stays on your Mac
There's a free plan with a weekly quota, no card required. For the Product Hunt community, code PRODUCTHUNT gives 50% off Pro for 3 months.
I'd love to hear where you type the most, and what would make you talk instead.
@serxo1 Hold a key, talk, text lands at the cursor in the tone you pick. That's a clean pitch. I spend half my day on calls and then rewriting notes afterwards, so the meeting minutes bit caught my eye more than the dictation. Does it handle two people talking over each other reasonably, or is it best with one clear speaker? Free plan with no card is a nice touch.
@upasana_sahu HI Upasana, I'm happy that you like and the Loqui also goes well when people are speaking over each other
What differentiates Loqui from WhisprFlow?
@sam_voahbstGreat question! The biggest difference for me is that Loqui goes beyond dictation into meetings:
• Smart minutes: it pulls out action items, what was decided and what was done.
• Spaces: you group meetings by project or client, and each space keeps the full context of every meeting in it.
• Conflict detection: if you or someone else says something that contradicts what was agreed before, Loqui flags it.
• Speaker names: it identifies who said what.
On the dictation side, it also adapts the tone per app (casual in Slack, formal in email), and it can edit text you've already written just by telling it what to change. And we never store your audio.
Would love to hear what you think after trying it!
Does the tone setting work per app, for example casual in Slack and formal in Mail? And does dictation run on device or is audio sent to a server?
@yuv23639661Yes, dictation works per app. Audio is processed on our servers, but we never store it. If you opt in to help us improve Loqui, we keep only the transcript and the raw text, never the audio. We’re also exploring a local mode using Apple’s on-device speech for Mac users, and looking for a way to bring something similar to Windows.
The meeting side is what interests me more than plain dictation, especially conflict detection inside a Space. Flagging when something said later contradicts an earlier agreement sounds genuinely useful, but also easy to get wrong. How does it decide between a real contradiction and a decision that was simply updated? Does it show both statements side by side so the team can judge, or does it just flag it and move on?
@jcass99 Great question, and yeah, this is the part that’s easiest to get wrong, so I built it to be cautious. When a meeting in a Space gets its minutes, Loqui compares the new decisions against the decisions still active in that Space. It only flags something when following the new decision means you stop doing the old one. Refining it, adding detail, handling an exception, or deciding something new on the same topic don’t count. If it’s unsure, it stays quiet, because a false alarm teaches people to ignore every alert.
And it never decides on its own. You get both statements side by side, before and now, each with the excerpt where it was said, and you can jump to that exact moment in the recording. Then you pick: “mark as superseded” (the old one is retired and drops out of future checks) or “not a contradiction”. Nothing changes until someone chooses.
Names and technical terms are the bits dictation tends to mangle, so do they stay untouched when the tone setting cleans up a sentence?
@serxo1 Thanks Sergio, that makes sense. Does the dictionary carry over to the meeting transcripts too?
@neerajnathany not for now, but it's on my backlog
@serxo1 no worries, a backlog item is still good news! would meeting minutes or the dictation side get the first pass of that?
Tone control is useful, but I'd want names and technical terms to survive the cleanup unchanged. A polished sentence still needs to say exactly what I meant.