Grass Farming Isn't Equal, and the Gap Is Widening
There seems to be some misconception that grass farming is "set it and forget it." Download the browser extension, leave the PC on overnight, and magically mine some GRASS tokens. Unfortunately, this isn't true. Grass has over 8.5 million monthly active users on the network but fewer than 20 enterprise customers paying to run their bandwidth. 96% of operators will make zero. A small fraction of the remaining 4% will take almost all of the remaining rewards.
The Grass protocol recently announced a change to how the network rewards users. They have split rewards into Uptime Points and Network Points. Right now the grass token price is floating around $0.34. This is down approximately 91% from an all-time high of $3.89. The answer to "how much grass will a casual node operator earn?" has very little to do with whether the grass app is running and everything to do with technical choices and strategy decisions that most people don't think about.
Where Most Node Operators Go Wrong
One mistake people make is believing Uptime Points is the "primary metric." Uptime Points represent availability, or how often a node is up and reachable. Network Points represent bandwidth consumed by Grass' enterprise customers. Network Points are weighted higher than Uptime Points under the team's new rewards schema when distributing tokens. A node that is online 100% of the time but never selected to route traffic will earn only a fraction of the reward that a node gets for being selected to route tens of gigabytes of web scraping traffic.
Casual users will generally only run the extension on a single laptop while at their desk. Typically 8-12 hours a day. This means casual users slowly earn Uptime Points as the day progresses. Many casual users don't check to see if their node has been selected to receive any crawl work from Grass's roughly 20 enterprise customers.
The second mistake: ignoring connectivity. Grass' infrastructure routes web scraping requests through node operators. Nodes that demonstrate low-latency, high-uplink-capacity, and residential IPs are prioritized by the protocol. Running a node on a capped connection, free public wifi, or a smartphone hotspot means being deprioritized. How much grass crypto a node operator earns is directly proportional to how useful that node is to enterprise customers scraping petabytes of data to train AI models. This is not charity. This is a marketplace.
Hardware, Connection Specs, and Why Location Changes Everything
CPU and RAM literally do not matter for grass farming. The Grass extension/node combination is extremely light on resources. The limiting factor is the network card. Almost every high earner is running 100% wired Ethernet through an ISP that offers residential plans with dedicated (non-shared) upstream speeds. An $80/month residential fiber connection with 100Mbps upload is going to outperform a $60/month 500Mbps cable connection with 10Mbps upload because most tasks that pay out on Grass are upload-intensive.
The SERP API product currently in beta testing with Tier 1 SEO companies requires request profiles that favor steady upstream throughput as opposed to raw downstream bandwidth. It's less about needing a dedicated machine and more about needing dedicated internet. A Raspberry Pi on some uncrowded fiber is going to beat someone running Grass on a gaming PC hooked up to a cable box 7 days a week. Grass doesn't need CPU cycles, it needs a clean fast connection to the public internet. Users who monitored their Network Points over rolling 30-day windows discovered that going from wireless to Ethernet alone can net a 30%-50% increase in task allotments based on user-submitted data in the Grass Discord.
Bandwidth isn't weighted equally by Grass's routing algorithm either. Enterprise web scraping bots want links spread around the world to avoid rate-limiting by IP. A Lagos node might be more valuable than one in San Francisco because there are fewer residential nodes in Nigeria. That's location arbitrage. Grass incentivizes users located where node density is low. They earn more Network Points per hour because their bandwidth is helping to fill inadequate global coverage. Community research indicates nodes are heavily centralized in North America and Western Europe. Nodes in Southeast Asia, Latin America, and Africa perform more work for the same amount of uptime.
ISP choice matters too. Some ISPs use CGNAT which makes nodes difficult or impossible to reach. Others throttle clients after scraping patterns are detected. Getting a clean residential IP without traffic shaping is an advantage.
Running Multiple Nodes Without Burning Returns
High earners are typically running nodes on multiple devices. But diminishing returns come into effect way sooner than most think. Grass tracks IP address in the protocol. Running 5 nodes from the same residential IP address will not earn 5x the reward. Nodes deemed redundant on the same internet connection get progressively throttled and deprioritized. Enterprise clients require geo and IP diversity.
The strategy from high earners: spread nodes geographically and across multiple ISPs. Home, office, a relative's place. As long as each device connects to a different IP address and provides network and geographic diversity, it works. Some run theirs on VPS instances. There is a bias against datacenter IPs in Grass's protocol though, favoring residential. Nodes on datacenter IPs are flagged and deprioritized because they provide less utility as residential exit nodes for web crawling.
The general community consensus is around 3-5 optimally located nodes as the sweet spot because beyond that the marginal gains become administrative overhead. At the current grass coin rate, the economics don't justify managing more.
Revenue, Unlocks, and Whether the Math Sustains
Grass has fewer than 20 paying customers and enterprise deals signed through 2025 with major AI institutions. The company has started spending some revenue from data sales acquiring GRASS tokens from the open market (deflationary signal). Their $10 million bridge round led by Polychain Capital and Tribe Capital was token-weighted in GRASS purchases in October of 2025, a clear signal they think highly of the protocol's revenue runway.
Compared to Masa, another bandwidth DePIN project, Grass comes out ahead with its massive user base of over 3 million active node operators. Helium proved, though, that early network node growth is not always a good representation of future demand. Other DePIN models like AIOZ have found traction through different infrastructure approaches.
On the supply side, the picture is less optimistic. Approximately 45.8% of GRASS's total 1 billion supply is locked. Tokens are released each month to early investors and ecosystem contributors via linear vesting. In October of 2025, 181 million tokens were released into circulation (amounting to 58% of total supply) and grass price halved. In February 2026, another 55 million tokens entered circulation.
No matter how much value the highest earner can demand per Network Point, each point will be worth less and less tokens with each unlock period unless the enterprise data buyer side of the business is growing at a rate that can absorb dilution. Grass network's enterprise-ready SERP API and upcoming Android app release are their best short-term drivers for revenue growth.
Many people think the farming component makes grabbing some exposure and holding for the long term a viable passive strategy. This could not be more wrong. The highest earners on the network are optimizing infrastructure placement, running nodes on multiple networks, and timing token swaps around unlock periods. Passive operators are being outperformed by those who aren't, and at current prices, subsidizing them in the process. The Grass token thesis as an AI data infrastructure play may work through 2026, but the math only favors those who treat it like infrastructure operations, not a passive income side project.
