Acti Raises $5.3M to Turn the Keyboard Into an AI Execution Layer

Acti has raised $5.3 million in seed funding to build what it describes as the world’s first agentic keyboard, a bet that the next major AI interface may not be another standalone chatbot, but the keyboard people already use across email, messaging, social apps, and mobile workflows.

The round was led by BITKRAFT Ventures, the early-stage investment firm focused on gaming, interactive media, AI, and adjacent consumer technologies. BITKRAFT says it manages $1.05 billion in assets and backs companies across gaming, applied game mechanics, digital assets, and AI, making Acti a somewhat unusual but logical fit: the company is not building a game, but it is trying to create a more interactive, habit-forming consumer interface.

Moving AI Out of the App Silo

Most AI assistants still require users to stop what they are doing, open a separate app, provide context, copy the answer, and return to the original conversation. Acti’s premise is that this workflow is backwards. Instead of asking users to go to AI, the company wants AI to appear where intent already starts: inside the keyboard.

On its website, Acti defines an agentic keyboard as a mobile keyboard with an AI agent built into every text field. The goal is not just to improve grammar or rewrite a sentence, but to complete actions such as finding a link, drafting a reply, creating a meeting, summarizing text, translating a message, or running a workflow without switching apps.

That distinction matters. AI keyboards are not new, but most have focused on text enhancement. Acti is trying to push the category toward execution. The company’s interface is built around a “type, hold, preview, apply” flow, where a user writes what they need, holds the Acti Bar, reviews the suggested result, and then decides whether to insert or apply it.

Skill Keys and a Programmable Keyboard

The core product idea is built around Skill Keys. Any key can be programmed into a skill, allowing users to long-press a key to trigger a specific action. In the company’s examples, long-pressing a key could translate a message, create a Google Meet link, generate a reply, summarize text, send a map link, or run a custom workflow.

Acti’s Skill Builder is designed to make that process accessible without code. Users describe what they want in plain language, and the system assembles the prompt, tools, and output into a reusable keyboard Skill. Those Skills can remain private or be published to the Skill Hub, Acti’s marketplace-like layer for discovering and installing keyboard actions.

That gives Acti two connected bets. The first is a product bet: that users will want AI actions embedded directly inside the keyboard. The second is an ecosystem bet: that users and developers will create enough Skills to make the keyboard more useful over time. Acti says early users created more than 1,000 Skills in under two weeks, while the company’s public Skill Hub describes categories such as social, writing, developer, and productivity workflows.

The Context Layer Question

Acti’s larger ambition is to become a personal context layer for the AI agent era. CEO and founder Young Wang argues that today’s AI agents are limited because user context is scattered across apps. A keyboard, by contrast, sits across nearly all mobile text interactions.

That creates an important opportunity, but also a difficult trust problem. A keyboard is one of the most sensitive surfaces on a phone. Acti’s Google Play listing says the app does not record everything a user types, and that text is processed only when the user explicitly invokes the Acti Bar or a Skill Key. It also says API connections use encryption and that users control what they submit.

This will likely be central to whether Acti can move beyond early adopters. The more useful an agentic keyboard becomes, the more it may need access to context, apps, APIs, and repeated user behavior. The challenge is to make that context useful without making users feel that the keyboard is watching too much.

A Bet on the Keyboard as the Next AI Interface

The broader AI market has spent the past few years adding assistants into existing workflows. Acti is taking a different approach by trying to make the workflow itself programmable from the keyboard up.

That is a more ambitious interface bet than another chatbot wrapper. It also puts Acti in difficult territory, because Apple, Google, and major AI platforms all have reasons to own the same surface. But the keyboard remains one of the few mobile interfaces used across nearly every app, and that makes it a compelling place to test whether AI agents can become less like destinations and more like embedded infrastructure.

With $5.3 million in seed funding, Acti now has room to prove whether the keyboard can evolve from a passive input tool into an active execution layer. The outcome will depend not only on the quality of its AI actions, but on whether users trust a keyboard enough to let it become the place where personal context, app workflows, and everyday intent begin to converge.

Acti is available on iOS and Android, with the App Store listing describing it as a free iPhone productivity app and Google Play listing support for workflows involving tools such as Notion, LinkedIn, Google Meet, Google Calendar, live sports schedules, nearby restaurants, and custom workflows.