Author: Frank, PANews
In recent times, the hottest topic in the tech and startup circles isn’t a major company’s new model release but the nationwide “lobster farming” craze.
On one hand, the “lobster farming” boom has driven growth in related industries, with large model companies and cloud server providers raking in profits. On the other hand, how much real benefit Openclaw can actually bring to users remains a mystery. Although social media is filled with such myth stories, a closer look reveals most are virtual stories created to attract traffic.
Is lobster farming truly profitable? If so, who is actually making the money?
PANews has compiled data from TrustMRR, public cases on social media, project official websites, and cross-verified reports from multiple sources. To distinguish “verified real income” from online self-proclaimed myths, we excluded many rumors based solely on unverified claims or hearsay.
According to TrustMRR, the OpenClaw category page shows 153 recorded projects within this ecosystem, with total income of approximately $358,600 USD over the past 30 days. Analyzing the top 30 samples, their combined income accounts for 97.3% of the total. If we break down these projects and their underlying monetization logic by “industry value chain,” a stark truth emerges: the first to make money aren’t those using lobsters as products, but those helping others farm lobsters, teaching others how to do it, and those riding the hype with meme coins.
However, this isn’t the most authentic answer we seek. How exactly are those truly using Openclaw making money? To answer this, PANews has summarized five monetization strategies behind OpenClaw.
First: Selling “Shovels” and services — quick cash from exploiting “cognitive gaps”
The most discussed and financially impressive products in OpenClaw are often not specific applications but tools and one-click hosting services.
OpenClaw functions more like an infrastructure layer rather than a ready-to-use consumer product. It presents a high barrier for non-technical users. Once complexity exists, services will emerge.
Within TrustMRR’s approximately $350,000 USD in sample income over 30 days, projects related to “hosting deployment” and “one-click cloud hosting” alone contributed about $120,100 USD, accounting for 34.5% of the sample income.
A typical example is QuickClaw, which packages underlying capabilities into a mobile app priced at $3.99/week or $49.99/year. Its income over the past 30 days is about $8,782 USD.
In Chinese communities, this logic manifests more simply: “lobster proxy” services on second-hand platforms like Xianyu.
According to media reports, recently, “OpenClaw deployment proxy” services on Xianyu and Xiaohongshu have exploded in growth. Remote installation costs range from 100-300 RMB, while on-site services cost between 400-1000 RMB. During certain periods, the daily transaction volume of related services increased by 150% compared to the previous quarter.
The core of this logic is “earning from information and perception gaps.” Users are willing to pay to save 30 minutes of hassle, but this is a “window period” business. As official one-click deployment tools mature, the profit from pure proxy services will quickly diminish.
Second: Packaging AI expert personas — when “storytelling” becomes the most expensive product
If we go a step further, another more valuable layer in the OpenClaw ecosystem appears: not just deploying for you, but training your Agent. Among the top 30 samples in TrustMRR, projects related to templates, skill packs, and configurations contribute 26.4% of income.
One of the most credible and well-documented business cases at present is FelixCraft.
In early 2026, creator Nat Eliason launched an experiment. He named his OpenClaw robot “Felix,” invested $1,000 as startup capital, and let it build its own business. Within a week, Felix generated about $3,500 USD via Stripe. Additionally, the crypto community issued related meme tokens for this Agent, forwarding 60% of daily transaction fees, allowing Felix to earn tokens worth up to $100,000 USD in a week.
As a highly analyzable case, Felix has several features. First, Nat Eliason granted this AI high permissions, allowing it to autonomously post on Twitter and interact in communities. Before product launch, Eliason stated he had spent significant effort building the framework, including memory modules, security settings, workflow design, etc.
The reason for profitability, Eliason admits in a podcast interview, was an unexpected outcome. Essentially, Felix’s main revenue still comes from packaging his training process and results as a product for sale. The meme tokens’ gains are more about the story and traffic it creates.
Notably, the top-earning project in TrustMRR’s OpenClaw category, Claw Mart (a marketplace for Agent skills), was created by Felix. Its total revenue has reached $71,300 USD. The story of Felix—an Agent capable of creating projects and automating work—serves as the strongest endorsement for this product.
Felix’s success reveals a high-level path for OpenClaw commercialization: endowing Agents with continuous identity. When OpenClaw is packaged as a specific name (Felix), a sellable guide, a set of reusable skill packs, and a compelling “AI entrepreneurship” narrative, it transforms into a powerful personal brand with viral potential. However, the core obstacle isn’t AI itself but Eliason’s strong agent training skills and marketing ideas.
Third: Selling efficiency myths — using AI to work, monetizing through storytelling
Among all monetization paths, the most recognized is: replacing manual labor with OpenClaw, and profiting from the cost savings.
This has become a reality in content operations. Developer Oliver Henry named his Agent “Larry,” responsible for his TikTok account. Larry automatically calls large models to generate images, write titles, and upload drafts. Henry only spends about 60 seconds daily choosing background music and clicking publish.
Henry states that within five days, Larry’s videos exceeded 500,000 views, bringing in about $588 USD in revenue (from paid app recommendations in his videos). Additionally, Larry generated $4,000 USD through meme coin issuance. Interestingly, Henry’s tweet sharing this story has already garnered 7.1 million views, similar to Felix, where the story itself seems more commercially valuable than the Agent.
Fusheng, founder of猎豹移动, built a team called “30,000” with 8 Agents, achieving daily updates on their official account from a few articles per year to daily posts, and setting a record of over 1 million views on a post about how Agents work, attracting social attention.
This suggests that in content creation, whether Agent-generated content can go viral remains unproven. Most viral stories so far are about Agents making money or improving work efficiency. Stories about “lobsters” are currently the biggest topic in content creation.
Fourth: Deep industry customization — moving beyond tool competition to earn “service premiums”
If proxy services earn “entry barrier” profits, then extending this concept by packaging “lobsters” into personalized products is another level.
RoofClaw exemplifies this. TrustMRR shows it earned about $49,800 USD in the past 30 days, with total revenue reaching $1.8 million USD. It offers “personalized customization and delivery of a MacBook Air equipped with OpenClaw system,” meaning its business isn’t just pre-installing a lobster but encapsulating it within a MacBook, along with tailored services to train the lobster into a suitable Agent.
This type of service likely taps into the future real commercial needs of “lobsters.” Users probably don’t just want a ready-to-use lobster but a fully trained, customized one. Behind this demand is a deep service for Agents.
Simply put, we foresee many companies relying on Agents in the future, but how to train or “coach” these Agents will become an unavoidable necessity.
Fifth: On-chain transaction legends — the most tempting poisoned apple and traffic bait
On social media, the most viral stories about OpenClaw are always about wealth miracles.
Currently, one verifiable on-chain account is 0x8dxd on the prediction market Polymarket, which is a high-frequency trading bot. Many social media posts speculate that this bot relies on OpenClaw for high-frequency trading, but PANews’s analysis shows the actual controller behind this address has never published such claims. The stories claiming “OpenClaw designed an automated trading system earning $100,000/month” are just soft ads, mostly promoting their automated trading programs.
This case is listed as a warning: as previous PANews research confirms, Agents and high-frequency trading bots are not the same. People are often misled and fantasize about their mystery.
Final reflection: Those who teach you how to make money are the real winners
After analyzing the entire ecosystem, we notice a phenomenon more worth pondering than any single case: sharing “I made this much with OpenClaw” on social media is itself a very stable business.
When a post like “I earn 50,000 USD/month with OpenClaw” goes viral, traffic becomes a lure. The author naturally directs viewers to paid communities, consulting, or product links. “Showing off income” is the top of the customer acquisition funnel, and “money-making myths” are the strongest marketing material. This creates a perfect self-reinforcing cycle: selling success stories — attracting traffic — monetizing traffic — then sharing more secrets as a mentor — leveraging even greater influence.
Essentially, this has spawned a new business chain: bottom layer is proxy and infrastructure, middle layer is skill packs and workflow replacements, top layer is industry solutions and consulting. If you understand sales, marketing, and have traffic, OpenClaw can drastically reduce costs and amplify productivity.
Many are sharing how OpenClaw optimized workflows and enabled many conveniences, but it’s far from a secret to wealth. The “herd effect” it triggers is the core of this traffic story: when you desperately push through the crowd to the front, you find nothing there, and you are the one waiting.
(PS: This article was created without using “lobster”)
