Hold Fn, speak, and Megaphone types clean text into any Mac app. Apple’s SpeechAnalyzer transcribes as you talk, while on-device Foundation Models remove fillers, fix self-corrections, adapt to app context, and power inline voice commands. No account, API key, subscription, or server. Free, MIT-licensed, and native to Apple silicon Macs running macOS 26.
Hey Product Hunt 👋
A few days ago, I was scrolling Hacker News when I came across a benchmark of Apple’s new SpeechAnalyzer API.
The results caught me off guard: Apple’s on-device model was more accurate and roughly three times faster than Whisper Small in the benchmark.
That made me wonder why so many Mac dictation apps still need subscriptions and cloud-hosted transcription.
So I built Megaphone.
Hold Fn, speak naturally, and release. Megaphone types clean text directly into whatever app you are using. SpeechAnalyzer processes your audio while you talk, and Apple’s on-device Foundation Models clean up fillers, resolve self-corrections, and fix punctuation before the result lands.
A few other things it can do:
• Start with “Hey Megaphone” to generate or rewrite text directly at your cursor
• Adapt formatting and vocabulary to the active app
• Learn names, acronyms, and technical terms in a private Dictionary
• Run in multiple languages
• Use hold-to-talk, toggle recording, custom shortcuts, and voice macros
There is no Megaphone account, API key, subscription, cloud transcription service, or server receiving your recordings.
Megaphone is free, MIT-licensed, and built natively in Swift for Apple silicon Macs running macOS 26 Tahoe.
It is built on top of Zach Latta’s excellent FreeFlow project. I replaced its cloud transcription stack with Apple’s SpeechAnalyzer and moved the cleanup and intelligence layer onto Apple’s on-device models.
I would especially love feedback on three things:
1. Accuracy with names, accents, and technical vocabulary
2. Whether Smart Cleanup changes too much—or not enough
3. Whether Inline AI feels genuinely useful in daily work
The source is public, the issue tracker is open, and I am actively maintaining it. Thanks for giving Megaphone a try.
no account or server for a dictation app is genuinely rare, most of them want a login and a subscription before you even test the accuracy. the self-correction cleanup is the feature I'd actually use daily, saying a sentence wrong and just re-saying it instead of manually deleting the messed up part. does it handle a correction that comes a few sentences later, like if you catch a mistake after you've already moved on, or does it only fix corrections said right after the error
Free, open source, and fully on device is such a rare combo for dictation. Hold Fn and just talk is exactly the right interaction too. Does it type into anything, even a terminal prompt? Congrats on the launch!
Honest question: how is this different/better than Apple‘s own dictation feature in macOS?
Hey folks, adding to launch day excitement, I'm running an agent loop that looks at the best suggestions in the comments and raises PRs to integrate it into the app 🙌
Running everything on-device feels like a long-term advantage if the experience is just as good as cloud tools.
Curious—what convinces people to stick with Megaphone after the first week—privacy, speed, or something else?
since everything is local with no account or server, what happens to the private Dictionary if you use more than one Mac. does it stay per machine and you re-teach it your names and jargon on each one, or is there some sync path that doesn't involve a cloud account
100% on-device is the part that sold me — no cloud round-trip for dictation means it should work fine offline and nothing leaves the machine, which is rare for this category. What's the on-device model you're running for transcription, and how's accuracy holding up against something like Whisper cloud on technical/code-heavy speech? Also curious if it's Swift/native or if you're wrapping something cross-platform under the hood.
About Megaphone on Product Hunt
“Open-source Mac dictation App that's 100% on-device”
Megaphone launched on Product Hunt on July 23rd, 2026 and earned 102 upvotes and 17 comments, placing #18 on the daily leaderboard. Hold Fn, speak, and Megaphone types clean text into any Mac app. Apple’s SpeechAnalyzer transcribes as you talk, while on-device Foundation Models remove fillers, fix self-corrections, adapt to app context, and power inline voice commands. No account, API key, subscription, or server. Free, MIT-licensed, and native to Apple silicon Macs running macOS 26.
Megaphone was featured in Productivity (656.7k followers), Artificial Intelligence (474.3k followers), GitHub (41.3k followers), Apple (15.5k followers) and OpenAI Day (9 followers) on Product Hunt. Together, these topics include over 288.7k products, making this a competitive space to launch in.
Who hunted Megaphone?
Megaphone was hunted by Kuber Mehta. A “hunter” on Product Hunt is the community member who submits a product to the platform — uploading the images, the link, and tagging the makers behind it. Hunters typically write the first comment explaining why a product is worth attention, and their followers are notified the moment they post. Around 79% of featured launches on Product Hunt are self-hunted by their makers, but a well-known hunter still acts as a signal of quality to the rest of the community. See the full all-time top hunters leaderboard to discover who is shaping the Product Hunt ecosystem.
Want to see how Megaphone stacked up against nearby launches in real time? Check out the live launch dashboard for upvote speed charts, proximity comparisons, and more analytics.
Kuber Mehta