Type what you need. Hold Acti Bar. Acti understands your intent and brings back the right result, link, or action - right where you are. Use Acti for live sports schedules, nearby restaurants, Notion docs, LinkedIn profiles, Meet links, Calendar actions, and custom workflows - without leaving the conversation.








@byalexai My favourite hunter! :)
@busmark_w_nika My favorite discussion (forum) creator on Product Hunt! Аnd soon my favorite maker there! :)
@byalexai Hahaha, thank you! :D
Yes! Two of my favorite people on Product Hunt in one thread! :)
The custom-workflows piece is what separates this from the rewrite/reply keyboards - pulling a specific Notion doc or firing a Calendar action from inside any text field is the real unlock. Two setup questions: is each integration a one-time OAuth connect per service, or do I wire it up again per workflow? And when I hold the Acti Bar, is the intent parsing on-device or does the field's text get sent to a server first - a keyboard sees everything I type, so where that runs matters.
@noctis06
Thank you , you’ve captured exactly what we believe makes Acti different from traditional rewrite and reply keyboards.
For integrations, each service only needs to be connected once through OAuth. After authorization, the same connection can be reused across different Skills and workflows, so there is no need to reconnect it every time.
On privacy, Acti does not continuously collect or upload everything you type. Processing only begins when you explicitly trigger the Acti Bar. Only the text required for that specific request is processed securely, and we do not log users’ keystrokes.
We completely agree that this distinction matters for a keyboard product, and we’re continuing to make the permission and data flow as transparent as possible.
One unexpected thing is how often I use ACTI for retrieval instead of generation. I'm not asking it to write essays. I'm mostly using it to fetch or execute things quickly.
@nicole_h94 That's exactly what we were aiming for. 😊 Instead of just generating text, we wanted Acti to help you retrieve information and complete tasks right from your keyboard. It's great to hear it's becoming a tool you use to get things done, not just write. How's your experience so far?
@nicole_h94 Thanks for the support — here's Acti Code to unlock LIFETIME Premium: 369MPV
@nicole_h94 Thank you so much for your support. Yes, that’s exactly how it works — it makes it easy for you to get the information you need directly in the chat box!
@nicole_h94 Totally. Being able to fetch or execute things quickly can be really helpful across a lot of everyday scenarios.
@nicole_h94 Thanks! That's exactly what we're building for—helping you retrieve information and get things done quickly, not just generate text.
Congrats on the launch!
For the Skill Keys, when you connect something like Notion or Calendar, does Acti ask for OAuth permissions upfront or does it only request access when you actually trigger that key?
@abod_rehman Thanks~ Yes you can set up the OAuth permissions upfront so that later when you use it, you wont have to request access everytime
@abod_rehman Thanks! You can grant permissions either in advance or when you run a Skill—whichever you prefer. We'll only access the tools you've authorized, and only when you manually trigger the Skill.
@abod_rehman Thanks for the support — here's Acti Code to unlock LIFETIME Premium: 287HNQ
This is a really interesting product! Congrats on the launch! I'm curious how it manages to accurately identify user intent, and what happens when the recognition isn't precise?
@llma Thanks for the support — here's Acti Code to unlock LIFETIME Premium: 39CJA5
@llma Overall, the recognition is quite accurate. You can give it a try — when it’s not fully sure about your intent, it will offer a few possible options for you to choose from~
@llma Thanks for your feedback! We've built a dedicated model for intent recognition, and when a user's intent is ambiguous, Acti will ask follow-up questions to clarify it before taking action. Our goal is to make sure the AI understands what you mean before it does anything.
Wow, the retrieval-over-generation point is really interesting. Four words from a keyboard carry way less signal than a full chat prompt, so I guess the tricky part is guessing whether I want a search, a doc or a triggered action. When the intent is ambiguous, do you commit to one guess and let me correct? Congrats on another great launch!
@artstavenka1 Thank you! In many cases, the user has already selected a specific Skill, so both the user and Acti know the intended action. For example, if you trigger the Document Skill, Acti knows you're asking about a document rather than performing a web search.
@artstavenka1 yes, when the intent is ambiguous, we let you to choose!
@artstavenka1 Thanks for the support — here's Acti Code to unlock LIFETIME Premium: 0J1DYR
Why did you choose the keyboard as the surface for agents instead of building another AI app?
@shengkun_ye Because the keyboard is the one interface people use across almost every app. Instead of asking people to switch to another AI app every time they need help, we brought AI to the place they're already using. The goal is to reduce app switching and let people get things done without breaking their flow.
@shengkun_ye Thanks for the support — here's Acti Code to unlock LIFETIME Premium: 23R2ZA
@shengkun_ye Thanks for sharing your thoughts! We believe that in the AI era, text input is becoming the universal entry point for getting things done. That's why we chose the keyboard—it naturally works across virtually every app, allowing users to access Acti's AI capabilities seamlessly without breaking their flow or switching contexts.