AI is powerful.
Chat is the
bottleneck.
Select anything on your screen and have AI explain, rewrite or translate it right there, without leaving the app or re-explaining the context.
1 month free at launch for the waitlist
Your invite + rare updates. Unsubscribe anytime. Privacy
One translucent screen you call up from wherever you are, and everything below runs inside it.
Select anything on your screen and have AI explain, rewrite or translate it right there, without leaving the app or re-explaining the context.
One press: the overlay opens over your apps
Circle the first thing you want to use
Circle a second one: both land in the command
Say what to do with them
Enter runs it: the answer appears in the bar
⌘S keeps the command for next time
Take the answer out as a card on the glass
One command on the mouse, from whatever app you are in.
Your screen stays visible underneath. You never lose your place.
Point at what is on screen instead of describing it in words.
One translucent screen you call up from wherever you are, and everything below runs inside it.Not three features in three places. One translucent screen you call up from wherever you already are, and everything below runs inside it.
Pick up a card you kept earlier
It opens into your note
Select a sentence
Choose an action: Explain
The answer lands in place, whole
Pick a follow-up
The change lands, highlighted
Pick up what you set aside, wherever you work.
Explain, Summarize, Rewrite, Ask, or one of your own.
The answer lands in place. No tab switch, no re-explaining.
Edit the answer the way you would edit a drawing.
Once a card is open, its text is yours to shape: pull a sentence longer, turn its words around, drag a passage out and back, brush a tone on, merge two passages, and keep every pass in a layer you can hide. No prompt in between.
An open card: its text is yours to shape
Pull the edge: the sentence grows
Turn the handle: the words reorder
Drag a sentence off the page
Right-click: split the sentence in two
Drop it back where you want: it blends in
Pick a tool: Smudge
Brush a word: it is rephrased
Erase a word: the sentence heals itself
Number: singular to plural, the rest agrees
Tense: present to future, the rest agrees
Type a rough note in the margin
Repair fixes the grammar
Or just ask: an example is added
Pick the eyedropper
Copy the tone of your own writing
Brush it onto the answer
Turn the wheel: another tone
One slider: a touch more positive
Edit a pole: Playful
Drag a passage over, Subtract: no repeats
Drag another over, Unite: one sentence
Hide the edits layer: your original is back
Show it again
Reorder the layers: the original on top
Pull a sentence longer by its edge, turn its words around by a handle, drag a passage off the page and back in, split a sentence in two.
Smudge to rephrase, erase a word and the sentence heals, change number or tense and the rest agrees, repair what you typed fast, or just ask.
Copy the tone of your own writing and brush it on. Explore tones on a wheel. Merge two passages. Every pass sits in a layer you can hide or reorder.
Every line of the answer shows where it comes from.
Circle a claim, keep the study you trust, and check what GalaxyBrain understood before it writes. Each line of the answer stays linked to its passage, and the links follow when you reshape the result.
Circle the claim you want checked
Research finds the studies behind it
Keep the most cited one
Compare the claim with the study
It shows what it understood. Drop a question
The answer arrives whole, each line linked to a source
Hover a mark: the passage lights up
Reuse the answer: rewrite it as a plan
Pull the edge to shorten. Undo if you want
Pick a tone
Drop it with a small overlap: the texts join
Drop it right on the plan: one text instead of two
Keep the whole chain as one action
Circle the claim. Research brings the studies behind it, sorted by how often they are cited.
Before a word is written, you see the sources and questions it will use. Drop what you don’t want.
Each line carries a source mark. Hover it and the passage lights up where it came from.
Rewrite it as a plan, pull its edge to shorten it, set the tone, merge two cards, undo. The marks stay. Keep the whole chain as an action.
Fix what it understood before a word is written.
GalaxyBrain shows the steps it will take as a graph you can edit: fold, delete, rename, add a step, or just say it. Confirm, and it writes from the graph you checked. Fix a step after, and it writes again.
Open what it understood: the steps, as a graph
It arrives folded. Open the steps you want to check
Hover a step: its words light up in the request
Delete a step: the request follows
Rename a step. No prompt needed
Add a step yourself. No prompt needed
Or say it: two steps are added, in green
Confirm: it writes from the graph you checked
A fix after: edit the step, it writes again
The card you checked opens into the steps GalaxyBrain will take, folded to what matters.
Delete a step, rename one, add one where you want it, or type what to change. The request follows.
It writes from the graph you confirmed. A small fix later: edit the step, and it writes again.
Your thinking isn't linear. Why should your AI workflow be?
Why am I the one carrying context between my tools?
Why is my work cut up along app borders instead of along how I think?
I already worked out how I want this done. Why do I have to explain it again?
Get your invite to the private beta.
1 month free at launch for the waitlist
Your invite + rare updates. Unsubscribe anytime. Privacy
- 1You're on the list
One email, nothing else to fill in.
- 2A few optional questions
About how you use AI today. They shape the beta.
- 3Your invite, by email
When the beta opens to you, with 1 month free at launch.
Write the prompt once. Keep it as a command.
Your best prompts live in a note you copy-paste from. Here you never save one: the moment a prompt runs, it becomes a command: the same instruction, with an empty slot where the thing you pointed at used to be.Right now your best prompts live in a note you copy-paste from, and you rewrite the context around them every single time. In GalaxyBrain you never save one. The moment a prompt runs, it becomes a command: the same instruction, with an empty slot where the thing you pointed at used to be.
Click a saved command: it stays armed
Circle a paragraph: it runs on it
Circle another: no retyping
Esc when you are done
Click compare: two empty slots
Circle the first thing: slot A fills
Circle the second: slot B fills, it runs
The answer lands in the bar
Take it out as a card on the glass
Circle first, then click the command
It runs once and does not stay armed
Take the answer out as a card
Tap the command, then keep tapping things down the page. Each one gets it. Esc when you are done.Click the command, then keep clicking things down the page. Each one gets it. Esc when you are done.
The command shows what you have filled so far. The last tap runs it. You never press send.The command shows what you have filled so far. The last click runs it. You never press send.
Point first, tap the command second. It runs once and does not stay armed, so there's no mode to escape from.Point first, click the command second. It runs once and does not stay armed, so there's no mode to escape from.
You already worked out how you want AI to do this. The point is not to make you explain it again.
Before you join.
What is GalaxyBrain, and where does it run?
A desktop app you run on your computer, next to the apps you already use. It's in private beta, and invites go out by email.
What does it read, and what leaves my computer?
Your screen is read locally, on your computer. Only what you select is sent to the AI, and only when you run an action on it.
Do I need an API key or my own AI model?
No. The AI model is provided by GalaxyBrain, so there's nothing to set up.
What happens after I sign up?
You're on the list. A few optional questions follow, about how you use AI today. When the beta opens to you, your invite arrives by email.
Help shape GalaxyBrain.
We're looking for the first people who feel this problem every day.
1 month free at launch for the waitlist
Your invite + rare updates. Unsubscribe anytime. Privacy