Ambient AI
companion.
Meet Jellyberry — a screenless, ocean‑inspired AI companion. Talk to it, and it answers in a natural voice and in light.
Get Early AccessJust tap
and talk.
Tap the shell and speak. No wake word and no screen. It remembers what matters between conversations, and forgets when you ask it to.
Ambient
intelligence.
Ask a question, set a timer or an alarm, or think something through out loud. Real back-and-forth conversation, with no phone to pick up.
“Set a timer for ten minutes”
Deep work,
made visible.
A Pomodoro timer you read at a glance. The light drains through 25 minutes of focus, then turns green for your break.
“Start a 25-minute focus session”
Bring the
outside in.
Rain, ocean, rainforest or fire, with light that moves to the sound. For working, winding down, or falling asleep.
“Play rain sounds”
Breathe in.
Breathe out.
The whole shell swells and fades in a slow four-second rhythm, with a soft bell between stages. Follow the light.
“Start a meditation”
Nature inspired.
Built for calm.
Modelled on the sea gooseberry (Pleurobrachia pileus), a comb jelly whose rows of light have rippled through the oceans for hundreds of millions of years.
Designed, coded and built by the sea in Brighton, UK. Inside the bioplastic shell: 96 LEDs, a speaker, a microphone and a single touch pad.
Catch the
first wave.
Jellyberry is a working prototype. Join early access for a place in the first small‑batch run, early‑adopter pricing and progress updates.
Frequent
questions.
- What is Jellyberry?
- How is it different from Alexa?
- What about privacy?
- What are voices?
- What can Jellyberry help me with?
- When will it be available and how much will it cost?
- Does Jellyberry need the internet?
- Where is Jellyberry made?
What is Jellyberry?
Jellyberry is a small, screenless device for people who want intelligence in their lives without the constant pull of a display. You hold it in one hand: 96 colour‑changing LEDs for responsive light displays, a microphone, a speaker, custom electronics, and a microcontroller inside a bioplastic shell.
It connects to Google Gemini Live, which means it holds real, flowing conversations — not just command-and-response. Ask it anything, follow up, change the subject. It listens, thinks, and responds fast and naturally.
There are four main modes: an AI assistant, a Pomodoro focus timer, guided meditation and nature soundscapes. It can also play internet radio and glow as a lamp. More on the way.
How is it different from Alexa?
Most smart speakers are command-response — you ask, they answer (or sometimes fail to) and the conversation stops there.
Jellyberry runs on a conversational AI model, so it holds real back-and-forth conversations. You ask a maths question, follow up with atomic theory, switch into planning your day, and it stays with you.
It also expresses itself through ambient light rather than a glowing screen, so you get a calm, present object in the room instead of another display competing for your attention.
What about privacy?
Jellyberry uses a touch‑triggered “conversational window” system. You tap to wake it, talk, and if you don’t reply within about ten seconds it quietly goes idle and stops listening until you touch it again
This means it isn’t always listening like many home devices, but still allows for naturally paced conversation when you want to ask more.
Audio is processed via Google Gemini, but reusable “memory” (AI context) is stored in a separate data layer you control. You decide how much, or how little, from your chats is saved and available in future conversations.
What are voices?
A voice is the way your Jellyberry assistant sounds and responds — its tone, pacing, and style of conversation.
All voices use the same underlying Jellyberry brain and access the same knowledge, but each has a different interaction style. One is calm and minimal, another more chatty and exploratory. You choose the voice that feels most natural to you.
At launch Jellyberry will ship with distinct voices you swap with a simple command. In future we’re exploring a voice designer so you can create custom voices without changing how your data or memory works.
What can Jellyberry help me with?
Jellyberry helps with everyday thinking, focus, and relaxation through four main modes.
You get an AI assistant for natural conversation, questions, timers and alarms; a Pomodoro focus timer; guided breathing for meditation; and nature soundscapes (rain, ocean, rainforest, fire). It can also tell you the tide, moon phase and weather, play internet radio, and glow as a lamp.
More modes and integrations are in development.
When will it be available and how much will it cost?
Jellyberry is currently a working prototype, with a first small-batch production run in development.
Early access members get first dibs on hardware and early-adopter pricing when we’re ready to ship.
Reserve your spot at jellyberry.ai.
Does Jellyberry need the internet?
Jellyberry needs an internet connection for conversational AI and live data features like tides and weather. Right now that means Wi‑Fi; we’re working on using your phone’s data so you can take it with you
The AI model and long-term memory live in the cloud, but several modes still work without an active connection — like the Pomodoro focus timer, lava lamp animations, and simple lamp mode. Jellyberry still feels alive even when you’re offline.
Where is Jellyberry made?
Jellyberry is designed and built in Brighton, UK, by the sea. It has been in development and testing for over a year, and the first batch will be hand-assembled — a genuine small-batch, maker-built product.
There's still time to catch the first wave.