Manus AI, a Singapore-based startup, has launched Wide Research, a new feature that orchestrates more than 100 parallel AI agents to tackle complex, large-scale tasks at speed. The feature debuted on July 31, 2025, and is available immediately to Pro-tier subscribers, with Plus and Basic access rolling out soon.
As VentureBeat reports, this new capability allows users to conduct tasks like comparing 100 sneaker models or profiling hundreds of companies simultaneously, delivering structured spreadsheets or sortable matrices within minutes. The multi-agent approach contrasts with traditional single-agent systems from companies like OpenAI and Google, which rely on deep analysis performed sequentially. (source:Business Insider)
How Wide Research Works
Manus describes Wide Research as powered by a cluster of general-purpose agent instances—each session runs on its own virtual machine, enabling agents to act independently and collaboratively in parallel. Unlike rigid role-based agent systems, each subagent is fully capable of handling diverse tasks without a predefined role or hierarchy.
(source:WinBuzzer)
Real-World Use Cases
In demonstrations, Manus used Wide Research to:
- Analyze 100 sneaker variants, generating a sortable matrix listing pricing and features
- Create 50 distinct poster designs simultaneously, packaged into a downloadable ZIP file
The design offers flexibility and speed for research, creative workflows, and enterprise data analysis.
(source:eluxr.com)
Why It Matters for AI
- Massive parallel processing—Wide Research achieves scale by running dozens of agents concurrently
- Enterprise-ready output—Results delivered in spreadsheets or dashboards, ideal for professionals
- Shift from depth to breadth—Offers an alternative model to single-agent, long-form AI research tools
As reported by The Outpost, the launch positions Manus as a challenger to established players using a radically different architectural approach.(source:Vox+Manus)
Limitations & Outlook
Wide Research is currently available only to Pro users (priced at $199/month). Broader deployment across Plus and Basic tiers is forthcoming, but Manus has yet to publish detailed benchmark comparisons or usage metrics. Users and industry observers have noted concerns around compute cost, redundancy, and result coordination in agent swarms.
(source:ypredict.ai)
