Verdict: Try it. The free plan needs no card, and Pro costs US$8 a month. The one thing to accept before you install: your voice is processed in the cloud by Loqui and its providers, so nothing runs on-device.

Loqui is a dictation app for Mac (with a Windows version) that launched on Product Hunt on 26 October 2026. You hold the fn key, speak, and cleaned-up text lands wherever your cursor is: Gmail, Slack, Notion, Cursor, Word, Jira. Loqui says it understands more than 90 languages, and it adjusts the writing style to the app you are typing in.

It also records meetings and turns them into minutes. That second feature is what separates the two paid plans.

Loqui comes from a single named maker. The privacy policy names the data controller as Sergio Vieira Augusto, operating from São Paulo, Brazil, and the Product Hunt listing names him as the maker. The policy names no company entity. That matters less for a US$8 app than for one you would roll out to a team, but you should know who holds your data.

If you mostly write long-form pieces you edit heavily, the gain is smaller. Dictation helps most where the gap between "what I'd say" and "what I'd send" is a quick tidy-up.

Dictation with tone by app. Loqui's homepage lists the default styles: professional in email, conversational in chat and WhatsApp, technical in code editors, structured in documents. You speak loosely and the text arrives formatted. Tap the key twice for hands-free mode, and tap once more to insert the text.

Rework after the fact. History is searchable, and you can redo a past dictation in another tone. "Spectros" let you translate, summarise or turn a dictation into an email draft. "Seals" are saved snippets you call up by voice, such as your booking link or a standard reply. Teams can share a glossary so product names come out spelt correctly.

Meeting minutes. Loqui records your microphone and the system audio on separate tracks, separates speakers, and writes minutes split into decisions, actions, risks and open questions, with audio clips attached. Before recording starts, it reminds you to tell the other participants.

What happens to your audio. This is the part to read carefully. According to the privacy policy (version 1.7, effective 24 October 2026), audio goes to Loqui's server, which passes it to ElevenLabs for speech recognition. Formatting is done by OpenAI models, directly or through OpenRouter, and Anthropic and Google may receive the text as a fallback. Loqui says its own server does not keep the audio or text and discards it when the request ends. The policy also says these providers handle data under their own terms, and Loqui does not claim they delete it immediately. Your dictation history stays on your Mac, and spaces and documents are copied to your Loqui account only if you turn on sync.

From Loqui's pricing page:

An extra 10-hour meeting pack costs US$6 on either paid plan. The page says the hours never expire, though you need an active plan to use them. No card is needed to start on Free. The pricing page says a launch coupon applies to monthly plans only; on Product Hunt the maker says the code PRODUCTHUNT gives 50% off Pro for three months.

Which paid plan (illustrative maths, not Loqui's figures). Paying annually for Pro costs US$80 against US$96 for twelve monthly payments, a saving of US$16. On meetings: say you record about 8 hours a month. Pro covers 3, leaving 5. One US$6 pack covers two months of that overflow, so you pay roughly US$8 + US$3 = US$11 a month. Pro + Meetings would cost US$16. Because pack hours never expire, they work out at US$0.60 an hour, so the bigger plan only wins once you regularly record more than about 16 hours a month (3 included plus roughly 13 from packs, which is where the packs pass the US$8 difference).

A one-week test worth running:

Before you dictate anything sensitive, decide what you won't speak into it. Client health details, passwords, legal matters under privilege: anything you would not paste into a third-party AI tool, don't dictate. The same rule applies to any cloud dictation app.

macOS Dictation (built in). Turn it on under System Settings > Keyboard. It adds commas, full stops and question marks automatically in supported languages, and Keyboard settings show whether your dictation is processed on the device. It stops after 30 seconds of silence and does not rewrite your words into a tone. Pick it if you want zero extra software and plain transcription.

Wispr Flow. The closest like-for-like rival. Wispr Flow's pricing page lists a free tier of 2,000 words a week on desktop, and Pro at US$15 a month or US$12 a month billed annually. It runs on Mac, Windows, iPhone and Android, and cites SOC 2 Type II and ISO 27001 certifications. Pick it if you dictate on your phone too, or need compliance paperwork for clients.

Superwhisper. Superwhisper's documentation says the free plan includes unlimited dictation with local Whisper models, and "nothing you dictate leaves your device". Pro costs US$8.49 a month, US$84.99 a year or US$249.99 one-off, and adds cloud models, AI rewriting modes and meeting recording on macOS. Pick it if on-device processing is non-negotiable.

If you are weighing whether you need another AI tool at all, our take on Notion Auto and choosing the agent where your work already lives applies here too. And since Loqui runs on ElevenLabs for recognition, our piece on ElevenLabs and voice AI going mainstream covers the company behind that engine.

Try it. Loqui costs nothing to test, asks for no card, and its US$8 Pro plan undercuts Wispr Flow's monthly price by US$7 while adding meeting minutes. Tone-by-app formatting and redo-in-another-tone are useful for anyone who writes a lot of short messages. Go in knowing your audio is processed in the cloud by Loqui and its providers, so keep sensitive material out of it, or use Superwhisper's local models for that work.

From today's Daily Briefing →