Acti launches AI keyboard that performs tasks directly from your smartphone keyboard

Acti Brings AI Keyboard with Agentic Capabilities to Smartphones

Singapore-based startup Acti launched on Tuesday with a new approach to AI integration: an AI keyboard that transforms how users interact with their smartphones 1

1

. Rather than forcing users to switch between multiple apps to access AI assistance, the company embeds agentic capabilities directly into the smartphone keyboard interface available for both iOS and Android 2

2

.

Founder and CEO Young Wang, who previously spent a decade at Baidu growing its Facemoji Keyboard to over 300 million daily active users, identified a critical friction point in current AI adoption. "Today's AI agents are fundamentally limited because user context stays fragmented across separate apps," Wang told TechCrunch 1

1

. The Acti keyboard sits across all applications, creating a context layer that belongs to the user rather than any single platform.

How the AI-Powered Smartphone Keyboard Actually Works

The core innovation centers on the ActiBar, which replaces the traditional space bar 2

2

. Users can press it normally to type or hold it to trigger actions. If a friend asks where to eat nearby, Acti can drop in a local recommendation directly into the chat. When someone mentions a stock, the keyboard can share the live price right there in the conversation 1

1

.

Under the hood, the keyboard is powered by Google Gemini models, chosen for their balance of intelligence, speed, reliability, multilingual performance, and cost efficiency 1

1

. This selection enables one of Acti's standout features: custom Skills that work like programmable shortcuts. Users can assign actions to any key on the keyboard, such as holding N to summon a specific Notion document or holding L to pull up a LinkedIn profile 2

2

.

Building Custom Skills with Plain-Language Prompts

Users don't need coding knowledge to create Skills. Instead, they describe what they want using plain-language prompts, and Acti builds the functionality 1

1

. The app ships with built-in Skills like "T" for translating messages to another language by long-pressing the letter, or "C" to instantly share a meeting link.

Ahead of launch, early access testers built over 1,000 Skills in less than two weeks 1

1

. These Skills can remain private or be shared publicly to a Skills marketplace, where users can find pre-built options for accessing real-time World Cup data or Polymarket links. This marketplace could offer additional monetization opportunities beyond the planned subscription model that will provide more advanced AI models, higher daily usage limits, and premium features.

Local-First Architecture Prioritizes User Privacy

According to both sources, Acti adopts a local-first architecture, meaning users' personal context stays on their device by default 1

1

2

2

. The company emphasizes that the app does not access or store private messages, conversations, or personal context unless the user explicitly invokes a feature requiring external processing. This on-device processing approach addresses growing concerns about user privacy while still delivering powerful AI assistance.

Funding and Vision for Human-Computer Interaction

Acti secured $5.3 million in seed funding in a round led by BITKRAFT Ventures 1

1

. "We backed Acti because this team has a real shot at owning the next phase of human-computer interaction," said Jonathan Huang, Partner at BITKRAFT Ventures. The team includes CTO Mike Sun, who was the founding technical lead behind Yike Album, Baidu's cloud-photo platform that scaled to over 10 million daily active users, and CSO Junbo Yang, who joined from HashKey Capital where he led dozens of consumer investments.

Wang's vision stems from recognizing that text has evolved beyond something people simply type. "When LLMs arrived, I realized something fundamental had changed. Text was no longer just something people typed; it had become a carrier of intent. And in many everyday contexts, that intent can now be directly translated into action," Wang explained 1

1

. This shift represents a fundamental rethinking of the keyboard, which has remained largely unchanged for two decades despite being the interface users interact with most frequently 2

2

.

The launch signals a broader trend in AI adoption: rather than asking users to change their behavior by opening separate AI chatbots, successful AI integration may come from embedding agentic assistant capabilities into the interfaces people already use daily. For professionals juggling multiple communication channels and productivity tools, this approach could reduce context-switching friction and accelerate task completion. As the Skills marketplace grows and users discover novel applications, the keyboard's utility may extend well beyond its initial use cases, potentially establishing a new standard for mobile AI interaction.