Voice-to-text dictation app with local (Nvidia Parakeet/Whisper) and cloud models (BYOK). Privacy-first and available cross-platform.
JavaScript aianthropiccross-platformgeminigroqlinuxmacosnvidiaopen-sourceopenaiparakeetspeech-to-texttranscribewhisperwindows
import assistantPreview from "../../assets/onboarding-assistant-preview.webp";
/**
* The OpenWhispr Assistant illustration on the assistant-hotkey step.
*
* A single exported image rather than a DOM composition: the artwork (grass
* backdrop, prompt bubble, mail window, dictation pill) ships as one 2x export,
* so it matches Figma exactly and cannot drift.
*
* Tradeoff, deliberately accepted: the copy inside the mock-up is pixels now, so
* it stays English in every locale. It is illustrative chrome, not UI the user
* reads for meaning.
*
* This stays above the shortcut controls and outside the shrinking flex pool.
* The frame is a fixed band rather than the export's 16:9 (at full height the
* step scrolled) and crops a lightly enlarged image biased toward the top, so
* the spoken request and mail window stay in view with no export edge showing.
*/
export default function AssistantHotkeyPreview() {
return (
<div className="mx-auto mt-3 h-56 w-full max-w-[30rem] shrink-0 overflow-hidden rounded-2xl">
<img
src={assistantPreview}
alt=""
aria-hidden="true"
width={561}
height={318}
decoding="async"
draggable={false}
className="h-full w-full scale-105 select-none object-cover object-[50%_35%]"
/>
</div>
);
}