🎤 Confident Spending

Whisper Tiny — separate voice test v131

This is a speech recognition experiment, not the live app. Press Load, wait for the model, then record a short command such as “monthly amount”.

Ready to load. First load may be large and slow on mobile.

Words recognised

Nothing yet.

Privacy: audio is processed locally by the model in your browser after downloading model files. No audio is intentionally uploaded by this page. Internet is required to fetch software/model on first use. Safari may pause the microphone when the page loses focus. This test does not yet play Jennifer's recordings.