● AI features require a Gemini API Key

AI you control
Use your data
Set the guardrails

Push to talk, in any language you speak. Search the web, query your data, jot down a quick thought, send an email — no distractions, full control: you set the limits, guardrails and data sources. Quick to set up, easy to audit with option to log every request for your review.

Free Shipping US and Canada - Assembled WISP32 watch: ESP32-S3, 2.06in AMOLED touch dev. kit, 3.7 V Li-Po Battery + 16GB microSD included. Open-source firmware (GPL-3.0)

WISP32 watch showing the clock face WISP32 watch showing the AI Assistant page

Specifications

Brain
ESP32-S3 Dual CoreReprogrammable Board
Display
2.06" AMOLED16.7 million color touch
Audio
Speaker + mic arrayFar-field capture
Storage
microSD cardAES-256 encryption
Communications
WiFi + BLE2.4GHz band

What it does

  • Discover, with you in control — you choose the sources, topics and limits
  • Searches your files automatically — just mention a subject and it finds the answer across your data
  • Continuous tracking — option to save text or audio of every AI answer for later review
  • Accelerates your learning — get information about new products and procedures while you work
  • Helps you remember what's next — spoken reminders, routines and next steps
  • A companion for kids' schooling — explanations and hints, guided by your guardrails
  • No distractions when you search — answers with no ads, videos or feeds
  • Ask instead of searching — get familiar info without opening an app on a tablet or phone

Screenshots

Clock face showing live weather data

Weather API

The clock face doubles as a live weather glance — current and feels-like temp, daily high/low, humidity, wind, pressure, rain chance, conditions, and the next sunrise or sunset, refreshes automatically.

AI Assistant screen confirming a file is open and ready for questions

Voice Assistant

Hold the button, ask a question, get a spoken answer — powered by Gemini's realtime voice models. Two modes, switched from a row on the page itself: Voice speaks the answer, Dual speaks it and shows the transcript on screen as it arrives.

Translator screen with input and output languages

Live Translator

Hold to speak in one language, release to hear the translation spoken back in another. Over 60 languages, curated to exactly the ones you use. Can save translated audio automatically.

Transcribe screen ready to record a voice note

Voice Notes

Tap, talk, done — transcribes your note to text on the spot, automatically names and saves the file, and can email it straight to a saved contact. Keep the text, the audio, or both.

Settings page 1 with volume, brightness, theme, and Device Lock controls

On-Device Settings

Six pages of settings live right on the screen — colour theme, brightness, AI model and voice, reasoning effort, auto-sleep, and more. Everything tap-to-cycle, no web login or reflash needed.

Save Modes

Choose to save AI Assistant and Translator mode output as text, audio, or both. Translator mode automatically saves the translation audio, ready to replay anytime from the device's web server.

Click to enlarge

QR Codes

Press and hold the clock face to pull up a QR code for the device's web server IP — scan to connect instantly. Step through your own saved QR codes too, added from the web configuration page — great for sharing your contact information.

Device Lock PIN keypad after a wrong attempt, showing retries left

Device Lock

Lock AI Assistant, Translator, Voice Notes, and the QR code page behind an on-device toggle — with an optional PIN keypad and a lockout timer after wrong guesses, so the assistant cannot be opened by accident, or by someone else.

QR code, enlarged

Questions

Everything you were about to ask.

Documentation

Read the manual.

WISP32 User Manual
Opens in a new tab