Manus Funding Targets $4 Billion After Its Meta Deal Unravels

Manus is pursuing $500 million at a $4 billion valuation, despite recently unwinding its acquisition by Meta and returning to independent operations. The reported Manus funding talks would give the AI agent company twice the valuation attached to its separation from Meta.

That increase makes this more than another large AI financing. Manus is asking investors to treat regulatory intervention as a detour, rather than evidence that its international structure remains vulnerable.

The round has not closed, and its terms can still change. Yet the proposed valuation reveals how quickly the market has reassessed Manus since Meta agreed to buy the company in December 2025.

Meta once offered Manus distribution, infrastructure, and a place inside one of the world’s largest consumer technology companies. Manus must now prove that independence creates more value than that integration promised.

The Manus funding round would reset its value

Manus is negotiating its first major financing since its separation from Meta, with reported terms that would value the company at roughly $4 billion.

The company is discussing a $500 million round, according to the initial funding talks. The report cites anonymous sources familiar with the private negotiations.

IDG Capital, Boyu Capital, and battery manufacturer Contemporary Amperex Technology are reportedly among the prospective investors. Existing backers Tencent, HSG, and ZhenFund may also participate.

Neither a final investor group nor completed financing has been announced publicly. That distinction matters because private funding terms often move before closing.

A $4 billion valuation would nevertheless establish a striking new reference point. Manus reportedly returned to its founders and previous investors through a transaction valued near $2 billion.

The proposed round would double that figure within a short period. It would also stand far above the roughly $500 million valuation reported for Manus in April 2025.

Benchmark led that earlier $75 million investment, according to reporting on the company’s previous financing. The progression would therefore move Manus from $500 million to $4 billion in about 17 months.

Those figures do not describe identical transactions. A venture round, strategic acquisition, investor buyback, and prospective public listing can assign different values to the same business.

Meta could justify paying for strategic control, technical talent, and integration rights. New investors must instead value Manus as an independent company with its own operating costs and expansion plan.

The reported $500 million would provide more than a valuation headline. It would give Manus capital to rebuild operational capacity after separating from Meta.

Manus may need to fund computing, product development, distribution, compliance, and hiring without relying on Meta’s systems. It must also reassure customers whose accounts were affected by the separation.

The company is reportedly considering a restructuring that would support a future Hong Kong initial public offering. That process would subject its ownership, finances, and risk disclosures to greater scrutiny.

An IPO remains a possible route, not a confirmed outcome. The immediate test is whether private investors accept the $4 billion argument before public markets examine it.

Independence now carries a measurable cost

The collapse of the Meta transaction turned independence from a startup ideal into an expensive operating requirement.

Meta announced its Manus acquisition on December 29, 2025. Financial terms were not officially disclosed, although reports placed the transaction above $2 billion.

At that time, Meta described Manus as a general-purpose agent serving millions of users and businesses. A general-purpose agent is software that plans and completes multistep tasks with limited supervision.

Manus offered subscriptions for research, coding, and other forms of computer-based work. That practical focus gave Meta something beyond another language model.

The acquisition also offered Manus access to Meta’s infrastructure and global product distribution. Manus CEO Xiao Hong said the combination would provide a stronger foundation without changing how the product operated.

Those expectations did not survive regulatory intervention. Chinese authorities reportedly ordered the companies to unwind the completed transaction amid national security and technology-transfer concerns.

The separation became operationally visible in August 2026. Manus told users that it would resume operating independently and make account changes required in certain jurisdictions.

The company’s service transition required some users to back up data generated after Meta’s acquisition. Specified data was then scheduled for deletion before eligible users could restore their records.

Manus said the action was not connected to a security incident. It attributed the process to regulatory requirements tied to its return to independent operations.

The episode showed that separating an AI company involves more than transferring shares. User data, subscriptions, internal access, technical systems, and regional obligations must also be disentangled.

That work creates direct costs and less visible commercial friction. Customers may hesitate when access, storage, or account rules change because of ownership disputes.

Developers and enterprise buyers face another concern. They need confidence that an agent will remain available while retaining stable access to projects, integrations, and stored information.

This matters more for agents than for simple chat interfaces. An agent can accumulate task histories, reusable workflows, research, and business context over time.

Moving that material can become difficult once a team depends on it. The separation showed why customers need export options and independent records for important AI-assisted work.

A personal knowledge base can reduce that dependency by keeping source material organized outside one agent’s execution environment. It does not remove platform risk, but it improves continuity.

The new Manus funding round must therefore cover two jobs. It must finance growth while repairing the operational confidence disrupted by the Meta separation.

Investors will want evidence that affected accounts returned after the transition. They will also examine whether new customer acquisition remained efficient during the ownership changes.

The difference between user registrations and durable usage becomes critical here. Viral interest can generate millions of accounts without producing predictable retention or margins.

Manus previously said it had surpassed $100 million in annual recurring revenue before the Meta transaction. Annual recurring revenue estimates subscription income over a year at the current run rate.

That figure helped explain Meta’s interest, but it does not disclose profitability. It also does not reveal churn, computing expenses, customer concentration, or revenue retained after the separation.

Fresh capital can give Manus time to answer those questions. It cannot answer them by itself.

The Meta reversal changed the valuation story

Manus is asking investors to value it above Meta’s reported purchase price after losing the strategic buyer that helped validate that price.

The central reversal begins with the original deal. Meta agreed to acquire Manus after the startup’s rapid rise during 2025.

Manus had launched its general-purpose agent earlier that year. Its demonstrations attracted attention because the system appeared able to research topics, write code, and produce completed outputs.

Meta was also under pressure to strengthen its position against OpenAI and Google. An operating agent product offered a faster route than waiting for internal research to become a dependable service.

The acquisition was expected to place Manus capabilities across Meta’s consumer and business products. That could have brought the agent within reach of Meta’s global user base.

The companies presented the deal as a path to scale. Manus would keep operating its product while drawing on the resources of a much larger owner.

The December acquisition also promised a clean ownership change. Meta said Chinese investors would no longer retain interests in Manus after the transaction.

Regulators viewed the arrangement differently. Reports described concerns about technology, engineering knowledge, and a Chinese-founded AI company moving under American control.

Manus had already relocated most of its staff to Singapore in 2025. The regulatory response showed that headquarters and corporate registration did not erase the company’s earlier ties.

That point reaches beyond Manus. Cross-border AI transactions can face scrutiny based on founders, financing, data, talent, or technology origins.

The failed acquisition has now produced an unusual valuation sequence. Manus was worth roughly $500 million in a venture round, more than $2 billion to Meta, and potentially $4 billion to new investors.

The simple interpretation is accelerating commercial value. Investors may believe Manus has established a scarce position in the market for autonomous work software.

A second interpretation is more complicated. The $4 billion figure may include the financing required to restore operations and prepare an eventual listing.

It may also reflect competition for access to an asset that regulators prevented an American buyer from controlling. Strategic scarcity can influence pricing without improving the underlying economics.

The proposed valuation therefore tests two different beliefs. One concerns the product’s growth, while the other concerns Manus as a strategically important AI company.

Those beliefs can support the same number for different reasons. They can also create conflict when investors eventually demand measurable operating results.

Meta no longer serves as the primary commercial backstop. Manus must support its valuation through independent revenue, customer retention, and sustainable delivery costs.

The company also loses a natural distribution channel. It cannot assume that integration into Meta AI, Facebook, Instagram, WhatsApp, or business products will supply demand.

That independence has one potential advantage. Manus can serve customers without forcing its roadmap into Meta’s consumer platforms or internal priorities.

It can also pursue partnerships that might have been awkward under Meta ownership. Enterprise software providers and other technology companies may prefer working with a neutral agent platform.

However, neutrality only matters if Manus can fund the required infrastructure and integrations. Competing independently against larger AI companies demands both capital and focus.

The Manus funding proposal is therefore a wager on regained freedom. The valuation assumes that freedom will produce more future value than Meta’s ownership could capture.

AI agent competition leaves little room for transition errors

Manus must execute its separation while competing against companies that already control models, developer platforms, enterprise relationships, or consumer distribution.

AI agents sit above foundation models and coordinate tools to complete a user’s objective. They can browse information, manipulate files, generate software, or organize structured results.

That product layer has attracted model developers and application startups alike. OpenAI, Google, Anthropic, Microsoft, and numerous independent companies are all pursuing agent-like workflows.

Their approaches differ, but the competitive pressure is similar. Customers want reliable task completion, predictable costs, data controls, and clear accountability when an automated action fails.

Manus gained attention by emphasizing complete outcomes instead of isolated answers. A user could assign research or coding work and receive an assembled result.

That experience can be compelling when it works. It can also expose weaknesses that a conversational assistant might hide.

An agent must maintain context across several steps. It must select appropriate tools, recognize errors, and decide when human approval is necessary.

Each additional step creates another failure point. Long tasks can consume substantial computing resources even when the final output remains unusable.

Independent agents also depend on external model providers, cloud services, websites, and software interfaces. Changes elsewhere can affect their quality or availability.

Meta ownership could have reduced some dependencies through internal infrastructure and model access. Independence returns those choices to Manus while returning their costs as well.

The reported investors may see an opportunity in that flexibility. Manus can select models or services according to quality, geography, and customer requirements.

It can also position itself as an orchestration layer, rather than competing directly to train the largest foundation model. Orchestration means coordinating models, tools, and task steps around an intended result.

That path still requires differentiation. Other companies can build similar interfaces on top of the same underlying models.

Manus needs proprietary advantages in task planning, tool use, reliability, workflow design, or customer data controls. A familiar brand and viral launch will not protect it indefinitely.

Enterprise adoption adds another challenge. Businesses evaluate procurement, security, auditability, support, and integration before allowing software to perform consequential work.

They may also restrict autonomous actions involving confidential files, customer records, code repositories, or financial systems. Those restrictions can narrow the tasks an agent completes without supervision.

Manus therefore faces two competing product demands. It must preserve the broad autonomy that attracted attention while adding controls that cautious organizations require.

The company’s geopolitical history can intensify those reviews. Buyers may ask where data is stored, which entities control the service, and which governments can influence operations.

Manus says its data is stored in the United States and Singapore after its return to independence. Customers will still need contract-level details for specific workloads.

Competitors tied to established cloud or productivity platforms can offer simpler purchasing paths. They may bundle agents with software that businesses already use.

Independent status gives Manus flexibility, but it removes the comfort of a large parent. The company must turn that flexibility into faster product improvements and clearer customer commitments.

For knowledge workers, the practical issue is continuity. Agent outputs matter most when they remain searchable, reviewable, and connected to the evidence that produced them.

A structured AI workflow can keep source materials and decisions available even when an execution tool changes. That separation helps teams evaluate new agents without surrendering their working history.

Manus can win customers by making complex work easier. It will retain them only if completed tasks remain dependable and portable.

A $4 billion valuation still needs harder evidence

The largest uncertainty is not whether investors want exposure to AI agents, but whether Manus can support its reported valuation with durable economics.

The proposed financing remains based on private discussions. Manus has not publicly confirmed a final closing, ownership structure, or complete investor list.

Reported participants may leave, reduce commitments, or negotiate different terms. The final round could therefore look materially different from the current proposal.

Valuation language also requires care. Reports do not always specify whether a figure is calculated before or after the new capital enters the company.

A $4 billion pre-money valuation would imply a higher post-money figure after adding $500 million. A $4 billion post-money valuation would give new investors a larger ownership share.

That distinction affects dilution and signals how aggressively the company is pricing the round. It should become clearer if Manus or an investor announces completed terms.

Revenue quality presents another open question. The previously reported annual recurring revenue figure suggested fast demand, but public reporting offers little detail about its composition.

Consumer subscriptions can grow quickly after a viral launch. They can also decline quickly if novelty fades or customers encounter inconsistent results.

Enterprise contracts usually move more slowly but can offer greater stability. They also create support, security, and compliance obligations that increase operating complexity.

Computing costs are especially important for autonomous agents. A multistep task can require repeated model calls, browsing sessions, code execution, and file processing.

Gross margin measures the revenue remaining after the direct cost of delivering a service. Without that information, investors cannot determine how efficiently Manus converts usage into value.

The separation from Meta complicates comparisons across reporting periods. Revenue earned under Meta may not map cleanly onto the newly independent company.

Some customers experienced subscription changes or temporary account disruption during the transition. Manus has not publicly disclosed retention following those events.

Nor has the company provided audited figures showing how many users regularly complete valuable tasks. “Millions of users” can describe reach without describing engagement.

A future Hong Kong IPO could force greater disclosure. Public offering documents normally provide financial statements, ownership details, business risks, and major dependencies.

However, considering an IPO does not guarantee that one will occur. Market conditions, regulatory reviews, and operating performance can all delay or cancel a listing.

Governance deserves equal attention. Manus has moved from venture backing to Meta ownership and back to founder-led independence within less than one year.

Existing investors reportedly participated in buying the company back. New investors will need to understand board control, voting rights, and the influence of major shareholders.

Tencent may emerge as a particularly important backer, according to current reports. Its final position and governance rights have not been publicly confirmed.

The company also remains exposed to relations between China and the United States. The failed Meta deal demonstrated that regulatory reach can cross formal corporate borders.

A future financing or listing may require Manus to explain how it separates sensitive technology, staff, intellectual property, and customer data across jurisdictions.

None of these uncertainties proves that $4 billion is unreasonable. They explain why the number should be treated as a proposal instead of an established measure of value.

Manus has already shown that investors and strategic buyers see meaningful potential in its product. The next stage requires evidence that this potential survives without Meta.

What to watch after the Manus funding talks

Three signals will show whether Manus has completed a genuine reset or merely replaced one transaction with another ambitious valuation.

The first signal is a confirmed financing announcement. The amount, valuation basis, lead investor, and governance terms will reveal how closely the completed deal matches current reports.

A round near $500 million with experienced institutional participation would strengthen the independence narrative. A delayed or reduced round would suggest resistance to the $4 billion target.

Investor composition matters as much as the headline amount. Strategic investors can contribute distribution or operating support, while financial investors usually focus more directly on eventual returns.

The second signal is customer retention after the separation. Manus should provide clearer evidence about recurring revenue, active paid customers, and usage following the August account transition.

Growth from the pre-separation period would strengthen the case that customers chose the product, rather than its connection to Meta. Weak retention would challenge the proposed valuation.

Product reliability should appear within that evidence. Manus needs to show that users repeatedly trust its agent with research, coding, and other multistep work.

New features alone will not settle that question. Useful indicators include repeat task completion, enterprise adoption, renewal rates, and improved delivery economics.

The third signal is formal movement toward a Hong Kong IPO. A restructuring, adviser appointment, or regulatory filing would turn a reported possibility into an observable process.

Offering documents would also expose information currently missing from the public debate. Investors could examine revenue, losses, ownership, customer concentration, and computing commitments.

A listing process may strengthen confidence if those disclosures support the growth story. It may weaken confidence if they reveal costly usage, unstable retention, or unresolved regulatory dependencies.

Competitor reactions will provide additional context, although they are not the primary test. Larger AI companies will keep adding agent capabilities to existing products.

Manus does not need to outspend every model provider. It needs to prove that its focused product produces outcomes customers will pay to repeat.

That is the real meaning of the latest Manus funding proposal. The company is not simply replacing the capital associated with its former owner.

It is trying to establish an independent valuation above the price supported by Meta’s strategic interest. That requires a stronger claim than rapid growth alone.

Manus must show that it can operate across borders, protect customer continuity, and absorb the cost of delivering autonomous work. It must do so without inherited distribution from Meta.

For developers and business buyers, the next decision is straightforward. Watch completed financing terms, post-separation retention, and any public filing before treating the $4 billion figure as settled.

The company’s return to independence has already happened. Its economic independence remains the part investors, customers, and competitors must now test.