DePIN Projects That Actually Make Money (Ignoring the Crypto Noise)

Let's get something out of the way upfront: most DePIN projects are garbage. Overengineered solutions looking for problems, propped up by token incentives that mask the absence of real demand. If you've developed a healthy skepticism of anything with "decentralized" in the name, you're not wrong to have it.

But here's the thing — buried underneath the hype, the vaporware, and the "tokenized everything" noise, there are Decentralized Physical Infrastructure Networks that are quietly generating real, measurable revenue. Not token rewards that are worth something only because speculators say so. Actual dollars, paid by real customers, for infrastructure they genuinely need.

On TBPN, Jordi Hays has taken to calling these "the cockroach DePIN projects" — they survived the crypto winter, they don't care about token price, and they'll still be here in 5 years because they solve real problems. This post is a guide to finding them, evaluating them, and understanding how much operators actually earn.

What DePIN Actually Is (In Plain English)

DePIN stands for Decentralized Physical Infrastructure Networks. Strip away the crypto jargon and here's the concept:

Instead of a single company building and owning all the infrastructure (data centers, cell towers, storage arrays), a network of individual participants contributes their own hardware — GPUs, bandwidth, storage, sensors — and gets paid when customers use it. The network coordinates supply and demand using blockchain-based protocols.

Think of it like Airbnb for infrastructure. Airbnb doesn't own hotels; it connects people with spare rooms to travelers. DePIN doesn't own data centers; it connects people with spare compute to companies that need it.

The blockchain component handles three things that are genuinely hard to do without it:

1. Trustless coordination: Proving that a node operator actually provided the service they claim to have provided
2. Automated payments: Paying thousands of small operators automatically based on usage, without a central payment processor taking a cut
3. Economic incentives: Bootstrapping supply before there's enough demand through token rewards (this is also where it goes wrong if demand never materializes)

The question isn't whether DePIN is a valid concept — it is. The question is which specific DePIN projects have crossed the chasm from "interesting idea subsidized by token rewards" to "sustainable business with real customers paying real money."

Category 1: Distributed GPU Rendering and AI Inference

This is the category with the strongest product-market fit right now, for one simple reason: there aren't enough GPUs. NVIDIA allocation remains constrained in 2026. Cloud GPU pricing from AWS, GCP, and Azure is extremely high. And the demand for AI inference is exploding.

Render Network

What it is: A decentralized GPU rendering network originally built for 3D rendering (VFX, architecture, gaming) that has expanded into AI inference. Node operators contribute NVIDIA GPUs, and customers submit rendering or inference jobs.

Real revenue indicators:

• Processing over 30 million frames per month across the network
• Customer base includes indie game studios, architectural visualization firms, and AI startups
• Network revenue (what customers actually pay, excluding token incentives): estimated $3-6M/month in early 2026

What operators earn:

• Single RTX 4090: $200-500/month depending on utilization (typically 40-65% utilization)
• Multi-GPU setup (4x RTX 4090): $800-2,000/month
• Enterprise-grade setup (8x A100): $3,000-8,000/month
• Hardware payback period: 6-14 months for consumer GPUs, 8-18 months for datacenter GPUs

Why it works: The demand is real and growing. Studios that can't get cloud GPU allocations or can't afford AWS pricing at scale have a genuine alternative. The rendering quality is verified on-chain, so customers know they're getting what they paid for.

Akash Network

What it is: A decentralized cloud computing marketplace. Think "AWS but the servers are owned by thousands of individual operators." Supports GPU and CPU workloads with a Kubernetes-based deployment system.

Real revenue indicators:

• GPU utilization across the network: 55-70% (a strong signal of real demand)
• Active deployments: 5,000+ concurrent workloads
• Growing enterprise usage: several AI companies use Akash for burst inference capacity

What operators earn:

• RTX 4090 node: $300-700/month
• A100 80GB node: $2,500-5,000/month
• H100 node: $5,000-12,000/month (but hardware cost is $30K+)
• CPU-only nodes: $50-200/month (lower demand, lower margins)

Why it works: The Kubernetes-based deployment means customers can use standard tools and workflows. You don't need "web3 expertise" to use Akash — a DevOps engineer who knows Kubernetes can deploy in an hour.

io.net

What it is: A GPU aggregation network that clusters distributed GPUs into virtual GPU farms. The key innovation: io.net makes scattered individual GPUs appear as a single, large GPU cluster to the customer.

Real revenue indicators:

• Network has aggregated 500,000+ GPUs (though active utilization is a fraction of that)
• Customer base heavily weighted toward AI inference (model serving) rather than training
• Partnerships with several AI startups that can't get cloud GPU allocations

What operators earn:

• RTX 3090/4090: $150-400/month (utilization is more variable than Render or Akash)
• Revenue heavily depends on geographic location (latency-sensitive customers prefer nodes in major metros)
• Token incentives still represent a significant portion of operator revenue, which is a yellow flag for sustainability

Caveat: io.net's reliance on token incentives is higher than Render or Akash. The ratio of real customer revenue to token subsidies should be watched carefully.

Category 2: Wireless Networks

The second category generating real revenue involves decentralized wireless infrastructure — micro-cell networks that offload traffic from traditional carriers.

Helium Mobile

What it is: Evolved from the original Helium IoT network (which struggled with demand) to Helium Mobile, a decentralized 5G/LTE network. Operators deploy small cell radios (indoor or outdoor) and earn from actual data traffic offloaded from carriers.

Real revenue indicators:

• Partnership with T-Mobile for traffic offloading
• Helium Mobile offers a $20/month unlimited plan to consumers, with real subscribers
• Network handles measurable data traffic in deployed areas (Miami, New York, Los Angeles, and expanding)

What operators earn:

• Indoor hotspot: $50-150/month in areas with consistent data traffic (dense urban)
• Outdoor 5G radio (FreedomFi or Baicells): $200-800/month in high-traffic locations
• Hardware cost: $250-500 for indoor, $2,000-8,000 for outdoor 5G
• Payback period: 3-8 months for indoor in good locations, 6-18 months for outdoor

Why it works: Carriers genuinely need traffic offloading. Mobile data consumption is growing 25-30% annually, and building macro cell towers is expensive and slow. Micro-cells deployed by individual operators fill gaps in coverage and add capacity where carriers need it most. The revenue comes from carrier payments for data offloaded — real money for real service.

XNET

What it is: A decentralized CBRS (Citizens Broadband Radio Service) network. Operators deploy CBRS small cells and provide wireless connectivity using licensed spectrum, which offers better performance than WiFi.

Real revenue indicators:

• Smaller network than Helium Mobile but with higher per-node revenue due to CBRS spectrum advantages
• Focused on commercial venues (stadiums, malls, office buildings) where connectivity demand is concentrated
• Revenue comes from venue owners and carriers who pay for offload capacity

What operators earn:

• CBRS node in a high-traffic commercial venue: $300-1,200/month
• Hardware cost: $1,500-4,000 per node
• Payback period: 4-12 months in good locations
• Location quality is everything — a node in a busy mall earns 10x what a node in a suburban office park earns

Category 3: Distributed Storage

The third category with real revenue is decentralized storage networks — distributed alternatives to AWS S3 and Google Cloud Storage.

Filecoin

What it is: The largest decentralized storage network. Storage providers commit hard drive capacity, and customers store data. The network uses cryptographic proofs (Proof of Replication, Proof of Spacetime) to verify that data is actually being stored.

Real revenue indicators:

• Over 20 exbibytes of storage capacity on the network
• Active storage deals with real customers including Internet Archive, GenRAIT (genomics), and several compliance-focused enterprises
• Growing demand for archival storage that's cheaper than AWS Glacier

What operators earn:

• Mid-sized storage provider (1 PB capacity): $500-2,000/month in storage fees
• Revenue varies enormously based on deal flow — providers who actively seek enterprise clients earn significantly more
• Hardware cost: $15,000-40,000 for a 1 PB setup
• Electricity and bandwidth costs: $200-800/month
• Margins are thin for small operators — this works best at scale

Arweave

What it is: Permanent storage — pay once, store forever. Unlike Filecoin's ongoing storage deals, Arweave charges a one-time fee and the data is replicated across the network permanently.

Real revenue indicators:

• Primary use cases: NFT metadata storage, archival of web content, scientific datasets, legal documents
• Growing adoption for AI training dataset preservation
• The "pay once" model appeals to organizations that need guaranteed data permanence

What operators earn:

• Mining rewards based on storage capacity and data served
• Typical operator: $200-800/month for a moderately sized node
• Revenue model is less straightforward than Filecoin — more dependent on network economics than direct customer payments

How to Evaluate Whether a DePIN Project Is Real

Here's a framework for separating the signal from the noise:

Green Flags (Real Business)

1. Identifiable customers paying with dollars: Can you name companies that use the network for production workloads? If the answer is "well, there are token incentives for using the network," that's not real demand.
2. Operator revenue exceeds token incentives: If operators earn more from actual customer payments than from token rewards, the network has crossed the sustainability threshold.
3. Utilization above 40%: If the network's resources are more than 40% utilized by paying customers, there's real demand. Below 20% and the network is mostly being subsidized.
4. Growing customer base independent of token price: Does the network gain customers when the token price drops? If yes, the product has standalone value. If customer activity tracks token price, it's speculation masquerading as demand.

Red Flags (Likely Vaporware)

1. No identifiable customers: "Our network has 50,000 nodes" means nothing if nobody is paying to use them.
2. Revenue is 80%+ token incentives: The network is paying operators from its treasury to create the illusion of activity. When the treasury runs out, operators leave.
3. Hardware pre-orders before network launch: Selling $500 "miners" to retail investors before the network is operational is a red flag. The revenue comes from hardware sales, not network usage.
4. No clear demand bottleneck: If the problem the network solves has plenty of cheap centralized solutions, decentralization adds cost and complexity without value.
5. Team focuses on token price over product metrics: If the project's social media is about price predictions instead of customer wins, run away.

The Operator's Guide: Getting Started

If you're interested in operating DePIN infrastructure, here's a practical starting point:

For GPU Operators (Lowest Barrier to Entry)

1. If you already have a gaming PC with an RTX 4090, start with Render Network or Akash
2. Expected setup time: 2-4 hours
3. Expected monthly revenue: $200-500 with a single GPU
4. Key considerations: electricity costs (a 4090 draws 300-450W under load = $30-60/month in electricity), internet bandwidth (100Mbps+ needed), and noise (you'll want the machine in a separate room)

For Wireless Operators

1. Start with a Helium Mobile indoor hotspot ($249-499)
2. Place it in a high-traffic location you control (business, retail space, dense apartment building)
3. Expected monthly revenue: $50-150 in good locations
4. Key consideration: location is 90% of your success — a perfectly placed indoor hotspot outearns a poorly placed outdoor radio

For Storage Operators

1. Start with Filecoin — the tooling is more mature and deal flow is better established
2. Minimum practical setup: 32GB RAM, 8-core CPU, 1TB NVMe cache, 100TB+ HDD storage
3. Expected monthly revenue: $200-800 depending on capacity and deal flow
4. Key consideration: this is the most capital-intensive and operationally complex option — not recommended as a first DePIN venture

The Bigger Picture

DePIN isn't going to replace AWS, T-Mobile, or Google Cloud Storage. That's not the point. The point is that decentralized infrastructure fills gaps that centralized providers can't or won't serve: the AI startup that needs 50 GPUs for a week but can't get cloud allocation, the carrier that needs micro-cell coverage in a neighborhood too small to justify a tower, the archive that needs permanent storage at 10% the cost of S3.

The projects that survive will be the ones where the infrastructure demand exists independently of the token. If the network would be useful even without a blockchain — if someone would build it as a traditional startup — then the decentralized version has a chance. If the only reason to participate is token speculation, the project dies when the music stops.

As John Coogan says on TBPN: "The best DePIN projects are the ones you can explain to someone who has never heard the word blockchain, and they still think it's a good idea." That's the filter. Use it ruthlessly.

While you're evaluating your next DePIN node setup, take a break and browse our TBPN hats and stickers — because even decentralized infrastructure operators need good merch. And if late-night node monitoring is your thing, a TBPN tumbler keeps your coffee hot through every epoch.

Frequently Asked Questions

How are DePIN earnings taxed?

DePIN earnings are taxed as ordinary income in most jurisdictions, regardless of whether you receive payment in tokens or fiat. In the US, token rewards are taxed at fair market value at the time of receipt (this is the IRS's position for mining and staking rewards, which DePIN earnings are analogous to). When you later sell the tokens, any price change is a capital gain or loss. Important: the fair market value at receipt establishes your cost basis. If you receive tokens worth $500 and later sell them for $300, you have a $200 capital loss that can offset other gains. Keep detailed records of every token receipt — date, amount, and fair market value. Tax software like CoinTracker or Koinly can help automate this. Consult a crypto-savvy CPA for your specific situation.

What happens if a DePIN network shuts down? Do I lose my hardware investment?

This is a real risk, and it's happened — several DePIN projects from 2021-2023 have shut down or become effectively inactive. The good news: for GPU-based networks (Render, Akash, io.net), your hardware retains its value because GPUs are useful for many things beyond DePIN — gaming, local AI inference, or joining a different network. For wireless hardware (Helium hotspots), the hardware is more network-specific and may have limited resale value if the network fails. For storage, the hard drives retain value. Mitigation: choose networks with strong demand signals (as described above), and prefer hardware that has alternative uses.

Can I run DePIN nodes from my home, or do I need a data center?

GPU and wireless nodes can absolutely run from home or a small office. GPU nodes need a decent internet connection (100Mbps+), reliable power, and ideally a room where fan noise won't bother anyone. Wireless nodes need to be in locations with foot traffic and data demand. Storage nodes become impractical at home beyond a few terabytes due to power, cooling, and internet bandwidth requirements — large storage operations typically use colocation facilities where you rent rack space ($50-200/month per rack unit). The sweet spot for home operators is 1-4 GPUs or a single wireless hotspot — manageable noise, power, and bandwidth.

Are DePIN earnings stable or volatile?

Earnings in dollar terms are moderately volatile — typically varying 20-40% month-to-month for GPU operators depending on demand cycles, network competition, and the types of jobs available. Wireless earnings tend to be more stable because data offloading demand is relatively consistent. If you receive payment in tokens and hold, the volatility is significantly higher because you're adding token price volatility on top of demand volatility. Most experienced operators convert tokens to stablecoins or fiat regularly (daily or weekly) to lock in earnings. Don't hold tokens hoping for appreciation unless you're comfortable with the speculation risk.