dictate-button.io is a micro service which enhances site text fields with dictation capabilities.
It’s a customizable, framework-agnostic <dictate-button /> web component, along with a script that injects it into site text fields and a backend API that transforms speech into text.
The backend currently uses local LLM models. It’s self-hosted and fully Dockerized.
The frontend is open source, so users can customize and adapt it to their environments:
👉 https://github.com/kkomelin/dictate-button
Here's what’s on my roadmap:
- Develop user accounts and dashboard
- Add optional filters (e.g., grammar correction)
- Add different inject scripts (e.g., all or no fields enhanced by default)
- Talk to users
- Talk to more users
- Talk to even more users 😅
If you’re interested in following my progress, join the waitlist:
👉 https://dictate-button.io/#join-waitlist
And of course, any feedback is greatly appreciated!