Hong Kong-based Animoca Brands will invest up to $10 million in selected early-stage projects building on its AI agent platform, known as “Minds by Animoca Brands” or Minds.

In a statement of Tuesday, Animoca Brands said the initiative targets teams with clear product strategies, strong execution capabilities, and scalable business models. It is open to applications across multiple sectors, including gaming, finance, productivity, and social platforms, provided the projects integrate Minds as a core component.

Minds is a platform for deploying persistent, always-on AI agents without requiring users to manage infrastructure such as servers or software. The system enables both developers and general users to create and control customizable AI agents designed to operate continuously and independently.

In addition to funding, selected projects will receive technical support, platform resources, and access to Animoca Brands’ broader ecosystem, which includes a portfolio of more than 600 Web3-related companies and projects. The investment allocation is not tied to a fixed timeline and will be distributed based on project quality.

The company positions the platform within the emerging “agentic web,” a concept referring to an evolution of the internet toward decentralized networks of autonomous AI agents capable of interacting, collaborating, and transacting on behalf of users.

According to the company, the Minds platform includes features such as persistent memory across sessions, collaborative capabilities between multiple AI agents, and simplified deployment through messaging tools without requiring advanced technical expertise.

Yat Siu, co-founder and executive chairman of Animoca Brands, said the company aims to support projects contributing to the development of AI-driven digital ecosystems. “Just as blockchain redefined digital ownership, agentic AI will redefine autonomy, unlocking new forms of creativity, coordination, and economic participation,” the executive added.

Agentic AI – lessons from the real world