Manus Launches Wide Research to Power AI-Driven Large-Scale Cloud Tasks

Manus has introduced Wide Research, a new feature designed to enhance its AI-powered cloud computing platform. Building on its earlier release of Manus, a general AI agent product, the company now enables users to execute complex, large-scale research and data tasks with ease. Manus originally set itself apart by allowing anyone, not just engineers, to access the power of cloud computing through simple natural language prompts. Each session runs on a dedicated virtual machine, enabling users to handle diverse tasks ranging from creating customized presentations to securely testing advanced open-source software. The goal has always been to make high-performance computing accessible to all. The new Wide Research feature takes this concept further by scaling compute resources for users by 100 times, thanks to optimized virtualization infrastructure and efficient agent architecture. With Wide Research, users can easily process information from hundreds of sources at once, enabling tasks such as analyzing Fortune 500 companies, comparing top MBA programs, or exploring emerging AI tools. Unlike traditional multi-agent systems that assign fixed roles to agents, Wide Research deploys fully capable general-purpose agents working collaboratively. This flexibility allows for dynamic, creative solutions unconstrained by rigid task formats or domains. Manus describes this advancement as not just a research tool but a foundation for future innovations in parallel processing and agent collaboration. Wide Research is launching today for Manus Pro users, with rollout to Plus and Basic plans to follow. The company sees this as a major step toward scaling AI agent capabilities and hints at larger projects on the horizon. By enabling powerful parallel processing and lowering the barriers to advanced cloud workflows, Manus continues its mission of democratizing access to cutting-edge computing power and reshaping how people use AI-driven tools for complex problem-solving.