Changelog

Made in v1.5 – v1.5.4

This was one of our most significant infrastructure upgrades.

⚡ New Speed Engine (Now Default)

• Our re-engineered dictation core is now officially stable. Short-form transcription is near-instant, and streaming output is significantly faster and smoother.

• Removed random filler expressions caused by streaming artifacts.

🏢 Enterprise-Ready Features

• SSO support for seamless enterprise authentication.

• Sync with enterprise-required privacy settings, including daily transcript clears and privacy lock enforcement. Live now (teams.willowvoice.com)

🔐 Reliability & Startup Improvements

• Willow now launches automatically at login — no UI interaction required.

• Authentication persists across restarts correctly (many edge cases covered)

• Pasteboard data no longer leaks into third-party apps like Raycast.

✨ Cleaner Output by Default

• Partial output tags (etc.) and em-dashes are now removed.

• Notifications for AirPod connections are clearer and more consistent.

This was all about taking Willow from “a tool you open” to “a system that’s just there.” Always on, always fast, and ready for enterprise scale.

— Lawrence Liu (CTO)

Made in v1.3.8 – v1.4.2

This update focused on laying a stronger foundation for Willow: making it more stable, privacy-conscious, and controllable in the background.

🔒 More Privacy & Control

• Delete all local transcripts with a single click in Settings > Privacy.

• Cancel or recover transcriptions instantly using the Escape key, complete with a visual toast confirmation.

⚙️ Better Stability

• Personalization settings now persist across restarts.

🎧 Audio Intelligence

• Audio Ducking is now live (Settings > Audio) — system audio automatically lowers when Willow is listening.

• Quiet mode audio processing has been improved for better accuracy in low-noise environments.

This release sets the stage for everything that’s coming — faster, more secure, and more reliable dictation.

— Lawrence Liu (CTO)

Made in v1.3.5 - v1.3.7.

We’ve made huge strides in both performance and personalization.

First, Willow is now noticeably faster and more stable across the board. We’ve continued improving short-form processing speed, and recent issues with Chrome-related crashes have been resolved. If you ran into frustrating slowdowns or app drops—those should be gone.

We also polished our Slack integration and personalization features. Slack messages now adapt better to your tone and style, and we’ve stopped insights from accidentally showing up from apps like Cursor and Windsurf. That was unexpected (and kind of weird)—thanks to the users who caught it.

As part of a broader privacy and control push:

• Auto-dictionary is now much less aggressive. Words won’t get added randomly anymore—there’s now a local cache system that filters for intent.

• You can cancel and recover transcriptions using the Escape key, with a visual toast in the top-right corner confirming the action.

• Your Profile page now shows your first name, last name, and email clearly.

• And in Settings > Privacy, you can now delete all local transcripts with one click.

We also think we resolved the issue where Willow would ask you to update twice in a row (but let us know if you still see it).

Finally, we’re still actively working on improving personalization accuracy—especially making sure that it works seamlessly without interfering with basic listening. Some users reported that personalization tweaks caused hiccups in standard dictation. That’s being addressed right now.

Thanks again to all the early users pushing this forward. Your feedback is what helps us move fast, fix things, and stay ahead!

Lawrence Liu (CTO)

Made in v1.3.3 - v1.3.5.2

We’re incredibly excited to announce something no other dictation tool has done before: true automatic personalization.

Willow now learns how you write, not just what you say. As you make edits to your transcriptions, Willow adapts to your tone, syntax, punctuation style, paragraphing, and more. Over time, it starts writing the way you write. Automatically.

This means:

• If you always capitalize certain terms, Willow will too.

• If you like dashes instead of commas, Willow adjusts.

• If you speak in longer, more structured paragraphs, Willow starts formatting accordingly.

It’s completely ambient. You don’t need to tweak settings or teach it manually. Just use Willow as usual, and it gets better each time you do.

✨ Quick tip: if you ever want to re-onboard your writing style, just click “Redo Personalization Setup” in your settings panel.

We’re still working hard to perfect the entire Personalization experience! Keep us in the loop how it’s working for you :)

Chat soon!

Lawrence (CTO)

Made in v1.2.15.2 - v1.2.15.5

You’ll feel it immediately. Willow is way faster now, especially for short-form clips. Instant processing, almost no delay, and way smoother transitions. Long-form improvements are in the works too (coming soon), but for most everyday use, speed should feel nearly instant.

We also fixed a bunch of issues along the way:

• Server-side encoding bugs that were slowing down uploads (v1.2.15.4)

• Memory leak that caused crashes during longer usage sessions (v1.2.15.3)

• Hotkeys randomly disabling or not registering (v1.2.15.3)

• Capitalization errors in smart insertions (v1.2.15.1)

• Layout problems with Bepó keyboard (v1.2.15.2)

As always, keep the bug reports coming in! Really appreciate all of you for being here :)

Best,

Lawrence Liu (CTO)

Made in v1.2.13 - v1.2.15

This release is a major upgrade to the way you interact with Willow Voice, especially if you’ve ever wanted more control over how dictation is triggered.

You can now set up multiple custom hotkeys, including one for hands-free mode and one for “Hey Willow” activation. This means you don’t have to rely on saying “Hey Willow” anymore—you can map the activation to a keypress and have full control over when dictation starts, even in noisy environments.

We’ve also removed the requirement to double-tap to enter hands-free mode. You can toggle it directly with your assigned hotkey, making the experience much smoother and more intuitive. And for those of you who love full customization, we’ve added support for F4 and F5 key bindings as well.

My Input: I'm personally a big fan of my set of key bindings, as shown in the key screenshot. Here's my explanation and how I'd recommend binding keys:

1. Fn for the default Mac keyboard works great (super reliable and easy to access).

2. On my external keyboard, I have bound Ctrl+Option as my way to default-start dictation. This makes it so my control and option key both work in isolation, but together they activate Willow.

3. Having the “Right Command” for Willow Assistant comes in handy whenever rewriting text. We've also revamped how Willow Assistant interacts with text, so it ends up being a lot less verbose, which will help with the complaints of sounding too much like ChatGPT.

For hands-free mode, I love adding the space combination to either Fn or Control. We override system-level shortcuts, so both of these are great!

Alongside these major features, we shipped a long list of bug fixes:

• Fixed crashes when Willow processed certain apps

• Made “Hey Willow” compatible with selected text (wasn’t working before)

• Solved issues where double-tapping led to out-of-sync recordings

• Improved the onboarding experience—extra steps now route you cleanly into the dictation app

• Fixed a high-impact bug introduced in 1.2.13, where quick key taps would leak into the next recording. That’s been fully resolved now.

• Added a clamshell diagnostic—if your Mac is closed or locked, we’ll toast and audibly notify you that Willow can’t access the mic

• Finally, we now ignore disabled event taps, which previously caused rare background crashes

This set of updates gives you way more flexibility and reliability, both in how you use Willow and in how it performs behind the scenes.

More coming soon!
— Lawrence (CTO)

Made in v1.2.9 - v1.2.12

We’ve rolled out a series of updates from focused mainly on stability, bug fixes, and quality-of-life improvements.

In v1.2.12, we addressed a couple of critical crash cases—one caused by audio effects triggering at the wrong time, and another from quickly double-tapping the hands-free hotkey.

Both are now resolved. We also made visual improvements to the notch waveform to smooth out janky transitions, and fixed an issue where the changelog wouldn’t appear after updates.

v1.2.11 introduced a proper in-app changelog, so you can now see exactly what’s new every time Willow updates. We also fixed “Launch at Login” (still being monitored in rare edge cases), restored clipboard and image recovery functionality, and added support for more niche custom keyboard layouts like Gallium.

v1.2.10 focused on compatibility with macOS Sonoma 14.1 and 14.2, which had introduced a silent error when reading audio files. This bug prevented word counts from increasing and broke downstream features like post-processing. That issue should now be completely resolved.

And in v1.2.9, we fixed an audio lag issue that stopped the UI from activating properly. This also fixed some bugs where transcriptions weren’t saving with the correct timestamps.

These updates may seem small on the surface, but they remove a ton of friction under the hood. And they’re helping us gear up for the bigger features on the roadmap.

Thanks for all the bug reports, and keep them coming!

Lawrence Liu (CTO)

Made in v1.2.7.17 - v1.2.8.2

Hi everyone, Lawrence here from Willow Voice 👋

I’m excited to share our newest batch of updates, fixes, and features. A huge thank you to all of you for using Willow, giving feedback, and helping shape what this product is becoming.

Here’s what’s new in the latest versions ⬇️

v1.2.8.2 — Annual Plans

You can now subscribe annually for a lower monthly cost, our most-requested billing feature is here. Thanks to everyone who asked for it (and waited patiently)!

v1.2.8.1 — Microphone Control, Upgraded Default Mic Isolation

We've decoupled Willow's default mic selection from your Mac’s system default. This means fewer surprises, and better control when switching devices like AirPods, external mics, etc.

v1.2.8 — Smarter Than Ever Auto-Dictionary: Full Release

Willow now learns from the way you speak. If it hears something repeatedly that it doesn't recognize, it'll save it to your dictionary. No more retyping the same odd names or jargon over and over again.

Dvorak Keyboard Support For our Dvorak users: thank you for your patience. Willow now plays nicely with your layout.

🔊 v1.2.7.17 — Better With AirPods Improved Audio Handling

We’ve rolled out a new audio stack that handles Bluetooth devices, especially AirPods, more gracefully. Less glitching, less delay, better recordings.

As always — we’re building Willow for people who want to think faster than they type. Thank you for believing in that mission with us 💜

Happy dictating,

Lawrence Liu (CTO)