DePIN turns hardware people already own into paid network infrastructure. A hotspot on a windowsill earns for the coverage it serves, and a spare hard drive earns for the files it stores. I went through CoinGecko's explainer on the sector, and the part worth arguing about is revenue. These networks pay their tokens out against actual use of a service, and the creator argues that makes the sector steadier than the rest of crypto when prices fall. Whether that holds depends on numbers that are harder to pin down than the pitch suggests.
What DePIN means
A DePIN is a network of physical equipment owned by many people instead of one company, coordinated by a blockchain and paid for in tokens. The name is short for decentralized physical infrastructure network. The hardware sits somewhere real, and whoever runs it gets paid when someone else uses it.
CoinGecko used a ride-sharing comparison to make the model click: anyone with a car can sign up as a driver to earn cash, except this version runs on a blockchain. The comparison shows where the money comes from. A rider pays, a driver delivers the trip, and a platform matches them. In a DePIN the same shape applies to storage, bandwidth or mapping data, with one difference: the payment and the record of who contributed what sit on a public chain, not in a company's database. That is what the creator means by cutting out costly middlemen.
Projects in this space cover Wi-Fi hotspots, solar-powered systems, cloud services and transport networks. Individuals deploy hardware and offer the resulting service to anyone who needs it.
Tokens do three jobs. They pay for the service, they reward the people providing it, and they can be staked to vote on decisions. Staking means locking tokens up for a period, and here the reward carries ownership and governance rights as well as a payment. That combination is what separates a DePIN token from a coupon. If the network does well, the token is meant to be worth more, and the people running the hardware are meant to have a say in how it is run.
The creator's case for why this beats the traditional model rests on motivation, transparency and cost. Providers are paid directly for useful work, the ledger is public, and the middleman's cut shrinks. I would treat those as claims to test one project at a time. Transparency is easy enough to check on a block explorer. Whether the service is cheaper depends on what you count and who paid for the hardware.
Physical networks and digital ones
DePIN splits into two families. Physical resource networks, PRNs, are location-based: providers contribute hardware for connectivity, mobility, energy and similar sectors, and the resource is tied to a place, sometimes fixed to it. Wireless, geospatial and sensor projects fall into this group.
Digital resource networks, DRNs, are not tied to a location. Providers contribute computing power, shared bandwidth or storage, and cloud storage, geographic mapping and data sharing projects sit here.
The line between the two is blurrier than the labels suggest. Hivemapper's maps come from dashcams on streets people actually drive, which is location-bound by any practical measure, yet the output is a data product you could serve from anywhere. I would not lean hard on the PRN and DRN split when judging a project. What matters more is whether anyone pays for the output.
The projects themselves fall into six categories: computing, wireless, energy, AI, services and sensing. Each has its own hardware costs and its own customers, and that list is the easiest way to find comparable projects when you start looking.
Four projects that put the model to work
Four examples span those categories. Here is what each asks of its participants, plus the market rank at the time. Ranks move, so treat that column as a snapshot rather than a judgement of quality.
| Project | What participants provide | Token | Rank |
|---|---|---|---|
| Filecoin | Storage space for other users' files | FIL | Not listed |
| Helium | Long-range wireless hotspots | HNT | #93 |
| Hivemapper | Dashcam footage from ordinary drives | HONEY | #342 |
| Fetch.ai | Autonomous AI agents on an open network | FET | #57 |
Filecoin is a peer-to-peer network that stores files. Users pay for storage and distribution, and people with spare disk space get paid for holding other users' files, in FIL. Think of it as a marketplace for disk space where the price comes from supply and demand rather than a hosting company's list.
Helium is a distributed network of hotspots that together create public, long-range wireless coverage. Participants are compensated in HNT, the network's own cryptocurrency. The hotspot is the physical part: someone buys one, plugs it in, and earns when devices use the coverage it provides.
Hivemapper builds maps of the world by having participants drive with a dashcam and contribute the images. When a contribution makes the map more useful, the driver gets HONEY tokens, and the same tokens can be redeemed for access to Hivemapper's map data. It is the clearest case in the group of a network whose supply side is literally the traffic.
Fetch.ai aims to widen access to AI through a permissionless network, where anyone can join without asking a gatekeeper for approval. Anyone can build and deploy autonomous AI agents that handle tasks such as optimizing DeFi trading services, managing transportation networks or running smart energy grids. FET pays for services and transaction fees.
All four pay contributors in their own token, so the real value of the reward depends on that token's market price.
The numbers behind the sector
A report by Messari and EV3 counted more than 650 DePIN projects across the six subsectors in 2023, with more than $20 billion in market capitalization and over $15 million in revenue. Market capitalization is the total value of a project's coins in circulation, so it measures what investors think the sector is worth, not what customers paid for the service.
Those last two numbers deserve a closer look. Twenty billion dollars of market value against fifteen million dollars of revenue works out at more than 1,300 times revenue, which is a striking ratio even for crypto. Market cap and revenue are not measured the same way, and a sector total is not a company valuation, so I would not read the ratio as a verdict. I would read it as a reason to check what is inside the revenue line before believing the growth story.
The market cap closed the first week of February 2024 at 13 billion, below the more than $20 billion cited for 2023. The two figures do not line up neatly and may measure different snapshots or different sets of projects. Go to the underlying report before quoting either number in a conversation about the sector's size.
Why utility revenue held up better
The resilience claim has a number behind it. The same report found that during the last market cycle, DePIN on-chain revenue fell between 20% and 60% peak to trough, while the rest of crypto fell between 70% and 90%. On-chain revenue means fees users actually paid to use the network, as opposed to the value of tokens handed out to providers. Peak to trough means from the highest point to the lowest.
The creator put it this way: "Because DePIN revenues are anchored in utility rather than speculation, the ecosystem is also more resilient when facing crypto market fluctuations."
The subsector figures show how wide the gap ran. Two of the categories are DePIN subsectors and four are the rest of the market.
| Sector | Revenue decline, peak to trough |
|---|---|
| Layer 1 networks | -76% |
| Decentralized exchanges | -71% |
| Money markets | -88% |
| NFTs | -90% |
| Compute (DePIN) | -57% |
| Wireless (DePIN) | -19% |
Wireless, the steadiest row, lost a fifth of its revenue, while money markets and NFTs lost close to nine tenths. Fees for storage or connectivity track how many people need storage or connectivity, and that number moves slowly. Trading fees track trading volume, which moves with the price.
I would add a caveat. Revenue that starts small can only fall so far, and a young subsector with a handful of paying customers will look stable next to exchange fees that swing with the market. The comparison also assumes revenue is counted the same way across categories, which is worth checking in the report. None of that makes the resilience claim wrong. It makes it one to hold loosely until you have seen the methodology.
What the creator says DePIN does better than a company
Horizontal scalability comes down to adding more machines to a network instead of making each machine bigger. A storage network can bring more disk online when demand rises and let nodes go quiet when it falls, so capacity tracks demand more closely. Anyone who has paid a cloud bill can see why that appeals.
Control works the same way. Infrastructure sits with a collective of individuals who commit their own resources instead of with one corporation, and each provider's influence is proportional to what they contribute.
Cost is the claim that needs the most scrutiny. The creator argued that because individuals contribute the resources, a DePIN spends little or nothing to bring a network to life, and services arrive at high quality for the lowest possible cost.
That last claim is where I hesitate. The cost does not disappear when a network runs on other people's hardware. It moves to the people who buy the antenna, the drive or the dashcam, and they stay only if the tokens they earn beat the electricity bill and the wear on the equipment. The quality claim rests on the same foundation, since providers have to be paid enough to care about reliability. Neither advantage is automatic. Both depend on the reward staying worth something.
What could slow DePIN down
The first problem is comprehension. In the creator's words: "DePIN as a concept could be hard to grasp, with daunting technical aspects that might put off any potential participants." Projects have to explain themselves to two audiences at once, the people who run hardware and the people who buy the service, and those audiences care about different things.
The second is scale. DePIN is still fairly new, and only a small percentage of enthusiasts and facility owners take part. A network with few providers gives poor coverage, and poor coverage attracts few users, so projects need a design that pulls in enough of both.
The third is the economics of the reward, and it is the hardest. Providers have to be paid at least what it costs to run their hardware, and the network has to stay profitable while doing it. That requirement runs straight into the token price. If the token falls far enough, a hotspot that paid for itself in March stops paying for itself in June, and providers switch off. Fixed costs in local currency against rewards in a volatile asset is an awkward match, and I have not seen a design that fully solves it.
Where the creator expected DePIN to go next
The creator pointed to three directions. AI came first: as calls for wider access to AI grow, more overlap between AI development and DePIN networks that supply computing power looks likely.
Zero-knowledge technology came second. A zero-knowledge proof lets someone prove a statement is true without revealing the underlying data, which is useful anywhere a network handles private information. The creator suggested it could improve privacy and security in data sharing, identity management and transaction privacy.
Gaming came third, as an incentive design rather than a product category. Gamified missions that reward players for taking part could bring in people who would not otherwise set up hardware.
What to check before you back a DePIN project
The story is appealing: real hardware selling a real service. The details decide whether it holds for one network. These are the questions I would ask first.
- Is there a paying customer base, or does most of the revenue come from token emissions paid to providers?
- What does a provider earn after hardware, electricity and bandwidth at the current token price?
- Does the token have to be used to access the service, or does it exist mainly to pay providers?
- Who sets the price of the service, and can governance change it?
What a provider earns per network and what the hardware costs to run have to come from each project's own documentation.
Two habits cover most of the practical risk. Check the site address before you connect a wallet, because a fake front end that drains wallets is the oldest trick in this market. Start small as well, with a single device or a small stake, so a wrong call costs you a lesson rather than a position. Buying hardware to run a node is a spending decision with no guaranteed return, and the token that pays you can fall while you are still paying off the equipment.
Who pays for the service, and how much of that payment is revenue rather than a token subsidy? A project that can answer both with a number is worth a closer look. One that answers with a market cap has not answered the question.
