Manus AI is back in the headlines, and this time it is about money, not a takeover. On October 8, 2026, the general AI agent maker said it raised more than $500 million. It is the first funding round since Chinese regulators forced Meta to abandon its deal to buy the company.

At Debugged Pro, our teams test new AI agents for client work almost every week. So I went through the CNBC report, the company's own Manus 2.0 launch post and other coverage of the round. Below is what happened and what the product does now. I also cover how agencies in India and the USA should test it.

The Manus website on a computer in Shanghai, January 2026. Photo: Bloomberg via Getty Images, as published by CNBC.

Manus AI Funding: The Key Facts

According to CNBC's report by Anniek Bao, Manus's parent company Butterfly Effect announced the round on Thursday. Here are the confirmed details:

Size: more than $500 million.
Lead investors: private equity firm Boyu Capital and venture investor IDG Capital.
Returning backers: Tencent, HSG and ZhenFund.
Valuation: not disclosed. Bloomberg reported last month that the round would roughly double the valuation to $4 billion.
Use of funds: also not disclosed, as TechNode noted in its write-up.

If the $4 billion figure holds, Manus would be China's most valuable AI agent maker. That is a big claim for a company that only launched in early 2025.

How Manus Got Here

The timeline is short but messy. Manus launched in China in early 2025. It then moved its staff to Singapore after US venture firm Benchmark backed it. In its December 2025 update, Manus said it had raised $75 million in a round led by Benchmark before launch.

That same month, Manus said it had crossed $100 million in annual recurring revenue just eight months after launch. Later that month, Meta announced a deal to buy the company for about $2 billion.

Then Beijing stepped in. China's National Development and Reform Commission said it would "prohibit foreign investment in the Manus project." By that point, Meta had already started folding the Manus team and technology into its own systems. The deal was unwound, and Manus went back to running on its own.

Why Investors Still Back Manus AI

Many people expected the Meta saga to scare off investors. It did not. Dan Wang, China director at Eurasia Group, told CNBC: "The fundraising shows that the short-term fallout of the Meta case has been contained and investors are willing to back Manus as an independent company."

CNBC also pointed out that interest in AI agent startups has held up. That is notable, because the base models these agents run on keep getting better and cheaper. A thin wrapper around someone else's model is easy to copy. Investors seem to believe Manus has built more than that.

My read is simple. Money flows to agent companies that ship products people pay for. Manus has a revenue track record, and it shipped a major update right before closing this round.

What Manus 2.0 and Cue Actually Do

On September 28, the company published its Manus 2.0 launch post. It calls 2.0 a new architecture, not a version bump. These are the parts that matter most for agency work.

Manus 2.0 overview graphic showing Automations, Video Editor, Game Dev, Remote Control, Studio, Cloud Computer and the Cue app — Manus 2.0 feature overview from the official launch post. Image: Manus.

Cascade: Less Waste per Task

Cascade is the new in-house execution system under Manus. It keeps a project light at the start and pulls in special tools only when the task needs them. In one tested setup, Manus says Cascade used 23.2% fewer tokens, finished tasks in 28.2% less time and cost 32% less to run than the old system.

Treat those as vendor numbers from one configuration. Still, cost per task is the number agencies care about most, so it is worth measuring on your own work.

Automations and Cloud Computer

Before 2.0, Manus could only start work at a scheduled time. Now a task can start when something happens in a connected service. Manus lists triggers such as a new email, a calendar event or a Slack message. A change in ad performance or a Notion update can also start a task.

Cloud Computer is a paid, dedicated environment for projects that need to stay online. Think of a game server or an automation that runs around the clock.

Manus Studio, Video Editor and Game Dev

The desktop app is now called Manus Studio. It handles documents, spreadsheets, PDFs, slides, websites, code, games and video. The Video Editor gives you a real timeline. You can swap a song or drop in your own product shot without regenerating the whole clip. Manus says it is built for short product ads of 30 to 60 seconds, tutorials and similar content.

Computer Use on Your Own Machine

Manus can now work on your computer during a connected, authorized session. It uses only the files, browser and apps you approve, and you can watch it live from your phone. This is powerful, and it is also the feature that needs the most care on work laptops.

Cue app by Manus shown on two phones with colorful mascot characters and a list of personal agents — Cue, the new standalone personal agent app from Manus. Image: Manus.

Cue: Agents With Their Own Identity

Cue is a separate app for personal agents. Each agent gets its own email address, phone number, wallet and computer. It can send messages, pay within a budget you set, take calls and leave you a summary. You can also put several agents in one group chat and have them pass work to each other.

Cue is in early access and free with an invite code. iOS support is waiting on App Store review. For the bigger picture, our post on personal AI agent standards and the Personal Agent Protocol covers the industry effort to make these agents work together.

What the Manus AI News Means for Agencies in India and the USA

Funding news on its own does not change your work on Monday. But this round tells us three useful things.

Manus is likely here to stay. Half a billion dollars buys a long runway. You can evaluate it as a serious tool, not a short-lived experiment.
The general agent race is now crowded. Meta pushed ahead with its own Muse agent in early September. Big platforms and well-funded startups now compete for the same users, which tends to push prices down and features up.
Vendor risk now includes geopolitics. CNBC described Manus as a cautionary tale for companies squeezed between regulators in Beijing and Washington. One ruling already changed who owns this company. That deserves a line in your risk register.

For US agencies, the question is less about features and more about client policy. Some clients, especially in finance, health and government, will ask where an AI vendor is based and who owns it. Have a clear answer ready.

For Indian agencies serving overseas clients, the same applies. If you handle personal data of Indian users, the Digital Personal Data Protection Act duties apply too. Check where Manus stores data before any client file goes in.

Where Manus Fits Next to Other Agents

Here is how I would map the new features to typical agency jobs:

Manus feature	Good agency use	Main thing to check
Manus Studio slides and docs	Pitch decks, research briefs, competitor summaries	Fact accuracy and source links
Video Editor	First-cut product ads and tutorial videos	Music and image rights
Automations	Ad performance alerts, weekly client reports	Which accounts it can read
Computer Use	Testing a web app while you work on something else	Device policy and approved apps
Cue agents	Internal scheduling and research helpers	Spending limits on the agent wallet

If your team already tested Meta's agent, our Meta Muse checklist for small businesses uses a similar approach you can reuse here.

A 30-Day Manus AI Pilot Plan

Do not roll a new agent out to every client at once. Here is the pilot we would run at Debugged Pro:

Week 1: Pick two internal tasks. Good picks are a market research brief and a short product video. Use no client data.
Week 1: Set a budget. Decide your spend cap up front and track cost per finished task, not cost per prompt.
Week 2: Compare against your current tools. Run the same task in the agent you already use. Note time saved, edits needed and errors found.
Week 3: Test one Automation. Connect a sandbox Slack channel or a test ad account. Never start with a live client account.
Week 3: Review data handling. Read the privacy policy and terms. Ask where data is stored and whether it is used for training.
Week 4: Decide. Write a one-page summary for leadership with costs, results and risks. Then choose to expand, pause or drop it.

Not sure which tasks are worth automating first? Our AI use case discovery service helps teams rank ideas by value and risk before they spend money on tools.

Red Flags That Should Pause Your Pilot

The agent asks for broad access to email or drives it does not need for the task.
You cannot get a clear answer on where data is stored.
Costs per task swing a lot from run to run with no clear reason.
A client contract bans third-party AI processing of their files.

Key Takeaways

Manus AI raised more than $500 million, led by Boyu Capital and IDG Capital, with Tencent, HSG and ZhenFund returning.
The valuation is not confirmed. Bloomberg reported about $4 billion, but Manus has not disclosed it.
Manus 2.0 adds real agency tools, including Automations, a timeline Video Editor and Computer Use.
Cue gives each agent its own email, phone number and wallet, so spending limits matter.
Treat ownership and data location as part of vendor review, especially for regulated clients in India and the USA.

About the author: Nitin Kumar is the Lead Project Coordinator at Debugged Pro. He manages web, mobile and AI delivery projects for clients in India and the USA. He also tests new AI tools with the team before they reach client work.