Knowii Voice AI

I’ve recently released the first versions of Knowii Voice AI, a Voice-to-Text app that runs 100% locally on your computer.

Press and hold a keyboard shortcut, speak, and words appear wherever your cursor is:

Check out the official Website to learn more: https://voice-ai.knowii.net

How it works

Underneath the cover, this app uses different AI models such as Parakeet (Nvidia) and Whisper (Open AI). Those run completely locally, meaning that you don’t even need access to the Internet to transcribe your voice.

And I have a long list of additional models I want to add support for. My first focus is privacy, but I will also offer the possibility of using cloud services (e.g., ElevenLabs, OpenAI, OpenRouter, etc). But that’s just the beginning…

Why?

This is a project that I’ve been working on for the past three weeks, with different goals in mind.

My very first goal was satisfying my curiosity about how to use local AI/machine learning models for Automatic Speech Recognition (ASR). I wanted to know more about the state-of-the-art in that area.

My second goal was diving back into some code, and learn something new (in this case, the Rust programming language and the Tauri cross-platform framework).

My third goal was trying a new experiment to see if I could actually build something people would want to buy and use.

Last but not least, I needed this for myself, and wanted to explore new ways to interact with my computer and AI. And I have a lot more ideas to connect/combine this with my knowledge management work, among other things.

Open Source Roots

In full transparency, I haven’t built Knowii Voice AI from scratch. I’ve actually forked an open source (MIT licensed) project called “Handy” (handy.computer) and I’ve built on top of the existing foundation, tweaking, bending, transforming and adding my own ideas.

I’m pretty convinced some people will view this negatively and judge, but the essence of the MIT license is about granting freedom to copy/reuse/modify/redistribute/sell/…

Please bear in mind that my goal is not to copy/paste whatever others do, but to explore and expand in other areas and in different ways.

Where is the source code?

Knowii Voice AI is closed source.

As much as I love Free and Open Source Software (and if you know me, you know that I do), I seriously doubt I could make this app Open Source right away and turn that into a sustainable source of revenue for my business. It’s sad, but it’s just the reality.

That being said, I’m not against making the code Open Source further down the road.

Business Model

The business model for this app is simple: pay $49 once, and get lifetime access. There’s no subscription, no license keys, no bullshit. You pay once, and you have it forever (including updates and support).

What’s Next

Since I’m building Knowii Voice AI as a product rather than as a hobby project, you can expect actual support and regular enhancements that go in the direction you need and want. The current roadmap is my vision about the next steps, but you and I will decide together where to go next.

Right now, Knowii Voice AI is “just” transcribing. That’s already quite cool, but I want to evolve it into a companion app that you can’t live without; understanding your current context, recognizing your intent, performing actions for you, and much more. Imagine saying “create my daily note for next Tuesday” and seeing it appear in front of your eyes. That the kind of feature I’m aiming for. Transcription is just level 1 in that game.

I’ve shared some of the ideas I have in mind for the future of Knowii Voice AI in the roadmap:

You can learn more about Knowii Voice AI on the official Website and in the official documentation. You can also take a look at the roadmap.

I hope that you’ll consider joining me on this new journey ❤️

A call for help 🙏

As you know, the hardest part when launching something new is attracting enough attention. I’m not really the type to spam everyone 100 times a day about what I’m building (maybe I should!), so that’s quite difficult for me.

All of you here in the Knowii community are the ones I want to keep helping and building for, and I know I can count on you.

I’m sure that many more people could benefit from this product, and from the other things I’m working on.

There are different ways for you to help me reach that goal, and to ensure I can keep working on valuable products. The most straightforward way to help is sharing the word about this product with your friends and colleagues: https://voice-ai.knowii.net

Let me know what you think. I'm always eager to get feedback and use it to improve my work.

Happy transcription!