The Decentralized Physical Infrastructure Network (DePIN) sector reached a combined market capitalization of $9-10 billion in April 2026, with leading networks generating approximately $150 million in verifiable on-chain revenue in January 2026 alone. The sector now spans three distinct verticals...
"It is decentralization theatre." — Sam Dare, Founder, Covenant AI (on exiting Bittensor, April 10, 2026)
The Decentralized Physical Infrastructure Network (DePIN) sector reached a combined market capitalization of $9-10 billion in April 2026, with leading networks generating approximately $150 million in verifiable on-chain revenue in January 2026 alone. The sector now spans three distinct verticals — GPU compute, wireless connectivity, and decentralized storage — each operating under fundamentally different economic models and facing divergent governance challenges.
This report compares the revenue structures, utilization metrics, and governance mechanisms across the three DePIN verticals using recent data from Aethir, Render, Akash (compute), Helium (wireless), and Filecoin (storage). The central finding: compute-focused DePIN protocols generate 10-40x more revenue than wireless or storage peers, but the Bittensor governance crisis of April 2026 exposed structural risks in token-mediated coordination that apply sector-wide.
DePIN's combined market capitalization sits at approximately $9.28 billion as of mid-April 2026, according to CoinGecko sector data. The World Economic Forum projects the broader decentralized infrastructure market could reach $3.5 trillion by 2028, though that estimate encompasses adjacent categories.
The sector divides into three operational verticals:
Each vertical monetizes differently. Compute networks charge per GPU-hour or per frame rendered. Wireless networks burn Data Credits pegged at $0.00001 per unit. Storage networks charge per gigabyte-month of sealed capacity. These structural differences produce widely divergent revenue profiles.
Across all three verticals, leading DePIN networks now trade at 10-25x revenue multiples, a material compression from the 1,000x+ ratios common during the 2021 cycle. According to Messari's sector analysis, this shift reflects a transition from speculative token pricing toward fundamental valuation.
GPU compute dominates DePIN revenue generation. Aethir reported annualized recurring revenue (ARR) of $166 million in Q3 2025, with Q3 alone recording $39.8 million — its highest quarterly figure and a 22% increase from Q2. Aethir offers enterprise-grade GPU computing at costs approximately 60-75% below AWS and Google Cloud equivalents, according to company disclosures.
Akash Network reports GPU utilization above 80% heading into 2026, with daily fee revenue hitting all-time highs exceeding $13,000 and total deployments growing 466% year-over-year to over 3.1 million in 2025. Akash's annualized recurring revenue reached $4.3 million — significant growth for the protocol but two orders of magnitude below Aethir.
Render Network processed over 68 million frames and generated $38 million in network revenue in January 2026. In April, governance proposal RNP-023 passed with 1.3 million yes votes against 15,500 no votes, approving the integration of Salad Network's approximately 60,000 consumer-grade GPUs as an exclusive subnet. All Salad payments and rewards will flow through the RENDER token, directly increasing on-chain transaction volume.
The compute vertical benefits from a structural tailwind: Bridgewater Associates projects major US technology companies will invest approximately $650 billion in AI infrastructure in 2026, and the global GPU market exceeded $125 billion in 2025 with projected 20%+ compound annual growth. DePIN compute networks position themselves as the cost-efficient overflow layer for this demand.
Helium remains the sole scaled DePIN wireless network. By Q3 2025, the network operated 115,750 hotspots (an 18% quarter-over-quarter increase) and had registered over 461,500 Helium Mobile accounts (up 48% QoQ). Average daily users reached 1.2 million, a 35% quarterly increase.
Revenue metrics tell a more modest story. Annualized revenue reached $18.3 million in 2025. Average daily Data Credit burns rose 196.6% quarter-over-quarter in Q3 2025, from $10,420 to $30,920. The Mobile Network accounted for 99.6% of that activity. Extrapolated annually, that pace yields roughly $11.3 million in DC burn revenue — a fraction of what compute networks generate.
Helium's partnership with T-Mobile and DISH Network provides a real distribution advantage. However, the economics face a structural constraint: Data Credits are fixed at $0.00001, meaning revenue scales linearly with data transfer volume. There is no pricing power. A Helium hotspot operator's return depends entirely on network-allocated HNT emissions, not organic fee revenue, creating subsidy-dependent economics.
Filecoin operates the largest decentralized storage network, with raw capacity of 7.6 exbibytes (EiB) and 2.1 EiB in secured data. Network utilization has reached approximately 31-32%, representing meaningful improvement from single-digit utilization two years prior but still leaving over two-thirds of network capacity idle.
Active storage from paid deals has stabilized at approximately 1,110 pebibytes (PiB). Paid deals increased 42% year-over-year according to blockchain analytics, indicating demand growth. However, protocol fee revenue tells a different story: Filecoin generated $180,700 in protocol fees over a recent 30-day period, according to Token Terminal — ranking eighth among Layer 1 blockchains but producing annual fee revenue of roughly $2.2 million.
The gap between Filecoin's massive infrastructure footprint and its fee revenue reflects a well-documented problem: storage provider economics depend heavily on block rewards (FIL emissions) rather than customer payments. The Filecoin community has identified AI data pipelines, DePIN integrations, and enterprise workloads as priority verticals for 2026, seeking to convert raw capacity into paid utilization.
On April 10, 2026, Covenant AI founder Sam Dare announced the project's exit from Bittensor, calling the network's governance "decentralization theatre." Covenant AI had trained a 72-billion-parameter AI model on the network and was among its most prominent subnet operators.
Dare alleged that Bittensor co-founder Jacob Steeves held unilateral control over subnet emissions, could revoke moderation rights, deprecate subnets, and use large token sales as "punitive" pressure — all without transparent community process.
The market reaction was immediate. TAO dropped from $337 to $253 in under six hours. Covenant AI liquidated approximately 37,000 TAO tokens (~$10.2 million), triggering over $10 million in additional long liquidations across exchanges.
Steeves responded by calling the exit a "deep betrayal" and proposed a new protocol feature: locked stake. Under this mechanism, subnet owners would lock tokens for defined periods to demonstrate long-term commitment, shifting from what Steeves described as "legal accountability to cryptographic accountability."
The Bittensor incident exposes a governance risk inherent across all DePIN verticals: networks depend on a small number of large infrastructure providers whose departure can destabilize both token price and network capacity. When a single subnet operator's exit triggers a 25% price decline and $10 million in cascading liquidations, the network's decentralization claims require scrutiny.
This risk is not unique to Bittensor. Helium's hotspot economics depend on foundation-set emission schedules. Filecoin's storage provider returns depend on protocol-determined block rewards. Render's subnet model concentrates capacity in approved partners. In each case, the gap between decentralized branding and operational reality creates governance surface area that has not been tested at scale until now.
| Protocol | Vertical | ARR (Latest) | Utilization | Revenue Source | |----------|----------|-------------|-------------|----------------| | Aethir | Compute | $166M (Q3 2025) | Enterprise contracts | GPU rental fees | | Render | Compute | ~$456M annualized (Jan 2026) | Per-frame rendering | RENDER burns | | Akash | Compute | $4.3M | ~80% GPU | Compute marketplace | | Helium | Wireless | $18.3M (2025) | 115K hotspots active | Data Credit burns | | Filecoin | Storage | ~$2.2M (fees) | ~31% capacity | Storage deal fees |
The data reveals a clear hierarchy. Compute protocols generate revenue at rates 10-100x higher than wireless or storage counterparts. This reflects both the higher unit economics of GPU time versus data transfer or storage, and the acute supply-demand imbalance in AI compute.
However, revenue concentration carries risk. Aethir's $166 million ARR derives primarily from enterprise contracts, meaning a small number of large customers drive the bulk of revenue. Render's January revenue spike was partially driven by one-time rendering demand. Neither pattern guarantees sustained growth.
Among the three verticals, Helium occupies the middle ground: lower absolute revenue than compute, but higher than storage, with a clearer path to organic demand through telecom partnerships. The constraint is pricing power — or rather, the complete absence of it.
DePIN's $9-10 billion market cap is supported by approximately $150 million in monthly on-chain revenue from leading networks, a material improvement over prior cycles where revenue was negligible relative to valuation.
GPU compute dominates DePIN economics. Aethir's $166 million ARR and Render's $38 million January revenue dwarf Helium's $18.3 million annual and Filecoin's $2.2 million in protocol fees.
The Bittensor governance crisis — triggered by Covenant AI's April 10 exit, a 25% TAO price collapse, and $10 million+ in liquidations — exposed governance concentration risk across all DePIN verticals.
Render's RNP-023 approval adds 60,000 Salad Network GPUs, but concentrates compute supply in a single subnet partner, replicating the operational dependencies that destabilized Bittensor.
Filecoin's 31% utilization rate and $180,700 in monthly protocol fees highlight the gap between infrastructure deployment and revenue generation that remains the sector's central challenge.
Sector-wide revenue multiples have compressed from 1,000x+ (2021) to 10-25x, reflecting a transition toward fundamental valuation — but sustained revenue growth has not yet been demonstrated through a full market cycle.
DePIN has progressed from conceptual framework to measurable economic activity. The sector generates real revenue from real customers: enterprises renting GPU capacity, mobile users consuming data credits, and organizations storing files on decentralized networks.
The divergence across verticals, however, is stark. Compute networks benefit from structural demand tailwinds in AI and generate revenue at enterprise scale. Wireless networks show organic user growth but lack pricing power. Storage networks command massive capacity but convert a fraction into paid utilization.
The Bittensor crisis adds a governance dimension that affects all three verticals. When protocol economics depend on token emissions, foundation-set parameters, and a small number of large operators, the gap between decentralized infrastructure and decentralized governance creates tangible risk. The sector's next phase will be defined not by whether it can attract capital or users, but by whether its governance mechanisms can survive the departure of key participants without cascading failure.