The Decentralized Physical Infrastructure Network (DePIN) sector now counts 650+ active projects and a combined market capitalization exceeding $16 billion, according to DePINscan data as of March 2026. The sector generated an estimated $72 million in on-chain revenue in FY2025 and attracted appr...
"DePIN is being forced into fundamentals. When token prices are flat, the only thing that matters is whether someone is actually paying for the service, and whether the network can sustain itself without subsidies. That shift is healthy." — Markus Levin, Co-founder, XYO Network
The Decentralized Physical Infrastructure Network (DePIN) sector now counts 650+ active projects and a combined market capitalization exceeding $16 billion, according to DePINscan data as of March 2026. The sector generated an estimated $72 million in on-chain revenue in FY2025 and attracted approximately $1 billion in private investment, predominantly at seed and Series A stages. A handful of projects — Aethir ($166 million ARR), Grass ($33 million ARR), Helium ($22–29 million ARR), and Akash ($4.3 million ARR) — account for the bulk of identifiable revenue.
Yet the sector's central tension remains unresolved: most DePIN tokens are down 94–99% from all-time highs, and the market trades at 10–25x revenue versus the 1,000x+ multiples of the 2021 cycle. The funding thesis is shifting from node-count growth to enterprise contracts, service-level agreements, and verifiable uptime. Whether this pivot produces self-sustaining economics — or merely replaces one subsidy model with another — is the defining question for DePIN in 2026.
DePIN encompasses blockchain-coordinated networks of physical hardware — GPU clusters, wireless hotspots, mapping dashcams, storage nodes, weather stations — where token incentives recruit operators and usage fees generate revenue. The sector's combined market capitalization reached $19 billion in September 2025 before settling to approximately $16 billion by March 2026, according to CoinGecko and DePINscan data.
The aggregate numbers mask significant concentration. According to Messari's State of DePIN 2025 report, the sector's estimated $72 million in FY2025 on-chain revenue is dominated by a small cohort of projects. Most of the 650+ networks generate negligible or zero protocol-level revenue. The World Economic Forum projects DePIN's total addressable market at $3.5 trillion by 2028, a figure that encompasses the infrastructure verticals DePIN targets — cloud compute, wireless, storage, mapping — rather than the blockchain-native market itself.
Private capital deployment into DePIN totaled approximately $1 billion in 2025. Escape Velocity closed a $62 million dedicated DePIN fund in December 2025, backed by investors including Marc Andreessen and Micky Malka of Ribbit Capital, according to Fortune. In March 2026, broader crypto VC deployment reached $3.1 billion for the month, with infrastructure and real-world-asset projects accounting for 65–70% of deployed capital, per CoinReporter data.
Four projects have established measurable revenue at scale:
Aethir leads the sector with $166 million ARR as of Q3 2025, operating a decentralized GPU compute network serving enterprise AI workloads. In October 2025, Predictive Oncology (NASDAQ: POAI) executed a $344 million private investment in ATH tokens to create the Strategic Compute Reserve (SCR), a pool of staked tokens designed to provision enterprise-grade GPU capacity. This represents the first instance of a NASDAQ-listed entity anchoring a DePIN project at scale. Aethir reported 1.4 billion compute hours delivered and claims cost advantages of approximately 70% versus AWS for equivalent workloads.
Grass generates $33 million in annualized revenue by monetizing unused internet bandwidth from 8.5 million monthly active users across 190 countries. The Solana-based project sells aggregated web data access to AI companies for training datasets. GRASS tokens surged 13% in mid-March 2026 following NVIDIA CEO Jensen Huang's remarks on agentic AI infrastructure demand.
Helium operates a Mobile Virtual Network Operator (MVNO) using T-Mobile's 5G backbone supplemented by community-owned hotspots. The network crossed 120,000 active subscribers in February 2026, generating $1.9 million in monthly revenue in December 2025 — an annualized run rate of $22.4 million. At 120,000 subscribers paying $20/month, implied monthly revenue is approximately $2.4 million ($28.8 million annualized). Data Credit burn doubled to $361,000/month by mid-2025, driven by carrier offload agreements with AT&T and Telefónica.
Akash Network reported $4.3 million in annual recurring revenue with 428% year-over-year usage growth and utilization rates above 80% heading into 2026, according to BlockEden data. Monthly compute volume reached $3.36 million.
The collision between decentralized compute and hyperscaler pricing is the sector's most consequential commercial front. AWS increased prices on key H200 GPU instances by approximately 15% effective January 4, 2026, without prior notice, according to Cloud Latitude analysis. AWS on-demand H100 pricing runs approximately $6.88/hour; Spheron's equivalent is $2.01/hour on-demand and $0.99/hour spot, per Spheron's published rate card.
The GPU rental market grew from $3.34 billion in 2023 to $7.38 billion in 2026 and is projected to expand another 28.7% in 2027, according to Thunder Compute's market analysis. Decentralized networks occupy a specific niche within this market: batch inference, short-duration training runs, and cost-sensitive workloads where 50–75% savings versus hyperscalers are material. For long-running, mission-critical training on clusters of thousands of GPUs, centralized providers retain structural advantages in network interconnect, reliability SLAs, and support contracts.
Render Network, which exceeded $2 billion in market capitalization, processes approximately 1.5 million frames monthly and has onboarded 600+ AI models. The project reported 35% of its 2025 output attributed to Hollywood studios and AI training clients, with a 40% increase in available compute power year-over-year.
Helium represents the longest-running DePIN experiment in telecom. The network pivoted from IoT (LoRaWAN) to cellular connectivity in 2023, and the MVNO model now accounts for approximately 90% of network revenue. The unit economics tell a specific story: 120,000 subscribers at $20/month generates roughly $28.8 million annualized, against which Helium pays wholesale rates to T-Mobile for 5G backhaul and distributes mining rewards to hotspot operators.
Total accounts registered exceeded 595,800 by December 2025, implying a conversion rate from sign-up to active subscription of approximately 20%. Helium introduced a freemium tier to drive adoption, a strategy that reduces average revenue per user but expands the addressable subscriber base. The network is licensing its hardware specifications to third-party manufacturers and pursuing carrier offload partnerships where telecom operators route traffic through community hotspots during peak congestion.
The telecom vertical faces a structural challenge: wireless infrastructure is heavily regulated, carrier relationships are complex, and the service must match consumer expectations set by incumbents spending tens of billions annually on network investment. Helium's path to sustainability requires either dramatic subscriber growth or enterprise offload contracts that generate higher-margin revenue.
The emergence of AI as a compute and data consumer has created a natural demand-side catalyst for DePIN networks that provision training data, bandwidth, and specialized datasets.
Grass's 8.5 million users sell access to public web data through a browser extension, creating what amounts to a distributed web-scraping network for AI training pipelines. The $33 million ARR figure positions Grass as the highest-revenue DePIN project on a per-user basis after Aethir. The project raised $10 million to expand its infrastructure, according to Ventureburn.
Hivemapper has mapped 29% of the world's roads within two years, capturing over 28 million kilometers of street-level imagery monthly — a rate Hivemapper claims outpaces Google Street View by a factor of five. Volkswagen's ADMT subsidiary uses Hivemapper's decentralized maps for its 2026 autonomous vehicle fleet. Hivemapper's Bee Maps subsidiary raised $32 million led by Pantera Capital. The project shifted from upfront hardware purchases to a $19/month subscription model, lowering the barrier for dashcam contributors.
GEODNET operates a decentralized network of GNSS reference stations delivering centimeter-level positioning accuracy for autonomous vehicles and precision agriculture. DIMO allows vehicle owners to monetize car data via token rewards, though specific revenue figures for both projects remain undisclosed.
Venture capital is repricing the DePIN sector along fundamentals rather than narrative. According to Messari analyst Dylan Bane, there were "no obvious catalysts to increase investment" in 2025, and DePIN projects should "prioritize achieving product-market fit on the demand side." The practical result: capital concentrates in projects demonstrating contracted enterprise revenue rather than node-count growth.
The Escape Velocity fund ($62 million) is the first dedicated DePIN vehicle of scale in 2026, targeting founders building infrastructure for telescopes, solar energy, and other physical networks. Broader Q1 2026 crypto VC deployment reached $5.9 billion year-to-date, with a $280 million growth round going to an Austin-based DePIN project, per CoinReporter.
An emerging financing model called "InfraFi" leverages the $175 billion outstanding stablecoin supply as collateral for DePIN infrastructure financing, potentially reducing dependence on token-denominated funding rounds. The model remains theoretical for most projects.
The foundational economic question for DePIN mirrors the broader blockchain economy's subsidy dependency documented across the sector. A peer-reviewed paper published in Frontiers in Blockchain (2025) states: "If a network cannot measure service and match it to demand, it becomes a subsidy machine. Emissions can bootstrap supply, but must converge toward revenue."
The numbers bear this out. The sector's $72 million in FY2025 on-chain revenue supports a $16 billion market capitalization — a ratio of approximately 222x. Even the leading projects trade at 10–25x revenue, which is compressed from historical crypto multiples but elevated relative to mature infrastructure businesses. Token emissions across DePIN networks — the primary mechanism for compensating hardware operators — constitute a subsidy flow that dwarfs protocol revenue for most projects.
The critical metric is the ratio of token emissions to fee revenue. For a DePIN network to be self-sustaining, usage fees paid by enterprise and consumer clients must eventually exceed the cost of token rewards distributed to operators. Currently, only Aethir and Grass approach this threshold, and both remain early-stage with concentrated client bases.
The "hardware trap" compounds the problem: DePIN projects face physical supply chains, regulatory exposure in multiple jurisdictions, and unit economics that differ fundamentally from pure software protocols. A GPU node operator must cover electricity, depreciation, and bandwidth costs regardless of network utilization.
DePIN occupies a specific position in the blockchain economy: it is one of the few sectors where tokens coordinate the provision of tangible, measurable services — compute cycles, wireless coverage, map data, storage capacity. That tangibility is both its advantage and its constraint. Unlike pure DeFi protocols, DePIN networks must cover real hardware costs, navigate physical supply chains, and meet enterprise service-level expectations.
The sector's trajectory in 2026 will be determined by whether the enterprise pivot produces contracted, recurring revenue at scale — or whether it results in a smaller number of projects capturing meaningful market share while the majority of the 650+ networks quietly wind down. The data so far suggests both outcomes simultaneously: a small cohort of leaders approaching commercial viability, surrounded by a long tail of projects still searching for product-market fit and sustained by token emissions rather than customer payments.