FinChip and AgentOn have agreed to merge and operate under the Finch brand, combining their AI-agent infrastructure, marketplaces and payment tools in a single platform. The companies also said they completed a pre-A funding round that values the combined business at $50 million, with Waterdrip Capital and Peer VC participating.
Finch is positioning the merger around a practical gap in the emerging agent economy: AI agents can perform tasks, but the systems that define their capabilities, find work, verify delivery and settle payments are often separate. The merged company plans to place those functions inside four connected markets covering agents, skills, human expertise and paid tasks.
The platform would allow a user or company to package an agent capability as a reusable skill, make it available in a marketplace, match it with a job and receive payment after the work is verified. Finch said task payments will use an escrow-style prefunding process, with automated settlement linked to delivery verification.
That structure places Finch closer to a labor and services marketplace for software agents than a conventional AI model provider. Its commercial test will be whether enterprises are willing to post repeat work and use a shared system for both agent discovery and payment, rather than treating AI tools as isolated software subscriptions.
Merger joins skill infrastructure with task matching
Before the transaction, FinChip concentrated on turning AI-agent capabilities into reusable and tradable assets. The company said it had built a network of more than 13,000 user agents and 18,600 skills, supported by over 100 specialists who distill industry knowledge into structured skills.
FinChip also reported completing more than 200 paid custom-development assignments, with cumulative task gross merchandise value exceeding $66,000. Those figures point to early experimentation with paid agent work, although they remain small relative to established enterprise software and outsourcing markets.
AgentOn approached the sector from the demand side. The company said its network matched agents with enterprise work, including longer-term roles and short-duration assignments. AgentOn reported providing work opportunities to more than 16,700 agents and paying more than $100,000 in wages.
Combining the two businesses could give Finch a more complete transaction loop. FinChip’s tools are intended to help define and package what an agent can do, while AgentOn’s marketplace is designed to route that capability toward buyers with specific work requirements. A unified platform could reduce the friction of moving from an agent profile to a paid assignment, particularly for tasks that require specialized workflows rather than a general-purpose chatbot.
Finch said proceeds from the pre-A round will be used to increase the supply of agents and skills across its four markets, expand enterprise demand and task flow, and advance adoption of its proposed on-chain skill standard.
The company previously raised funding from Longling Investment Group, ME Group and X INFINITY, according to its announcement.
Portable records sit at the center of Finch’s model
Finch plans to use the ERC-8004 standard for its agent identity and work-history layer. The company said the framework is designed to allow an agent to hold a portable identity and verifiable employment record across different platforms.
In practice, an identity and reputation layer could address a basic problem for agent marketplaces: a company hiring an agent needs some way to assess its previous work, capabilities and record of delivery. If that information remains inside one marketplace, an agent’s reputation may be difficult to carry elsewhere. A portable record could make performance history more useful across platforms, provided the underlying verification mechanisms are trusted by employers and marketplace operators.
Finch also said it has submitted an ERC-8338 proposal to the Ethereum Magicians community. The proposal is intended to define ownership and transaction rules for AI-agent skills.
The standard remains a proposal rather than an established industry rule. Its value would depend on adoption by developers, marketplaces and users that create, license or purchase agent capabilities. Without shared implementation, skills could remain tied to individual platforms even if the technical framework exists.
The company’s emphasis on standards also reflects a commercial choice. Finch is attempting to make agent skills behave more like identifiable digital assets, with ownership records and transaction rules that can persist beyond one vendor’s interface. That could appeal to developers seeking ways to monetize narrowly defined workflows, from customer-support procedures to data-processing tasks, without rebuilding distribution and payment systems for each use case.
Payment protocols are becoming a competitive layer
Finch’s launch comes as technology and payment companies are developing systems intended for transactions initiated by AI agents. Gartner has forecast that “machine customers” will participate in or influence around $30 trillion in procurement transactions by 2030, a projection that captures potential spending affected by automated purchasing systems rather than a measure of current agent-led commerce.
Visa, Mastercard and Google each introduced payment protocols for AI agents in 2025, according to the companies’ announcements. These initiatives focus on questions that become more urgent when software can initiate purchases or complete work: how an agent is authorized, how payment limits are set, how transactions are verified and who is accountable when an automated process fails.
Finch’s approach is narrower than building a universal payment rail. Its proposed settlement model is tied to work posted through its own marketplace, with funds prefunded and released after delivery verification. That could give companies more control over task budgets and reduce payment disputes, though it also makes the platform’s verification process central to user trust.
“NOW HIRING: AI” offers a public demonstration
Alongside the merger, Finch launched an interactive project called “NOW HIRING: AI,” designed to show how agents are deployed and how they perform tasks. The project gives the newly combined company a way to demonstrate its marketplace model before it has a lengthy public operating record under the Finch name.
The merger creates a business built around a specific proposition: AI agents need more than model access to become paid service providers. They need recognizable capabilities, work histories, job channels and settlement systems that employers can use with confidence. Finch now has to show that its four-market structure can generate repeat enterprise assignments and that its proposed identity and skill standards can gain use beyond its own platform.
Explore how AI agents transform real trading by trying Toobit’s AI copy trading tools in live markets.
Disclaimer: The content on this page is provided for general informational purposes only and does not represent the views or financial advice of Toobit. We make no guarantees regarding the accuracy or completeness of this information and shall not be held liable for any errors, omissions, or outcomes resulting from its use. Investing in digital assets involves risk; users should independently evaluate their financial situation and the risks involved. For further details, please consult our Terms of Service and Risk Disclosure.
