NEAR is going through a strategic reset. The project is no longer being framed mainly as a high-performance Layer 1 built around sharding, low fees, and fast confirmation. It is now trying to position itself as infrastructure for multi-chain transactions, cross-chain settlement, and AI agents, where wallets and software agents handle routing, swaps, and payments without requiring users to deal with blockchain complexity directly.
That shift already has some measurable traction. NEAR says NEAR Intents has processed more than $13 billion in cumulative cross-chain settlement. Some community estimates put the range at $20 billion to $26 billion across 30 to 35 blockchains. At the same time, NEAR has changed parts of its token model: the maximum annual inflation rate has fallen from 5% to about 2.5%, developer gas rebates have been removed, fee burns have expanded, and some product revenue has been used for token buybacks.
The problem is that rising activity has not yet translated cleanly into durable token demand. According to the revenue dashboard cited in the source article, NEAR-related products generated about $3.48 million in total fees in the 30 days before Sept. 1, 2026. After payments to solvers, partners, apps, and other participants, net protocol fees were about $757,500. Transaction volume has expanded quickly. The amount of value retained by the protocol is still much smaller.
From a general-purpose chain to a multi-chain transaction layer
NEAR began as a general smart contract platform aimed at consumer-facing applications. Its Nightshade architecture uses sharding to process network state with the goal of scaling capacity while keeping costs low. According to project disclosures, NEAR currently has block times of about 600 milliseconds and finality of roughly 1.2 seconds. In 2025, the network increased its shard count from six to nine, lifting theoretical throughput by around 50%.
Those performance metrics are respectable, but they are no longer enough to stand out on their own in the current smart contract market. Solana, Sui, Aptos, and Ethereum Layer 2 networks all offer fast or cheap transaction environments. For NEAR, staying focused on TPS and fees risks pushing it deeper into a crowded field where the products look increasingly similar.
That is why the project has moved toward chain abstraction. In simple terms, chain abstraction is meant to hide the underlying blockchain from the user. Instead of manually choosing a bridge, managing gas assets, and selecting a route, the user specifies an intended outcome and the system finds the network, liquidity, and execution path.
NEAR Intents is the centerpiece of that strategy. In a conventional cross-chain transaction, the user often has to choose the bridge, trading venue, and destination network. In an intents-based model, the user submits the desired result and solvers compete to find the best route across platforms and liquidity pools.
NEAR says Intents has already surpassed $13 billion in cumulative cross-chain settlement. Some community tracking puts the figure between $20 billion and $26 billion and says activity spans 30 to 35 blockchains. The article notes that different datasets may use different time windows and methodologies, but they still point to the same broad conclusion: NEAR Intents has reached a meaningful level of real transaction activity.
That changes NEAR’s potential market. A standalone blockchain usually grows by getting developers to deploy applications on-chain and then drawing in users and assets. If the intents model works, NEAR may participate in asset flows across many blockchains even when it does not host the applications themselves, and it may earn revenue from cross-chain transactions in the process.
There is a built-in tension, though. The smoother the user experience becomes, the easier it is for the underlying network to disappear from view. Users may benefit from execution powered by NEAR without knowing that their transactions touched NEAR at all, and without needing to hold the token directly.
NEAR is also not the only group trying to own that entry point. Wallets, aggregators, cross-chain protocols, and other intents networks are competing for order flow. Even if settlement through NEAR Intents keeps growing, much of the fee pool may still go to solvers, market makers, and partner applications that provide pricing and liquidity. How much revenue NEAR keeps is likely to matter more than how large the gross settlement number becomes.
AI adds a new story, but not yet a proven business line
AI is the other major leg of NEAR’s repositioning. The project is working on autonomous agents, confidential model execution, verifiable AI output, agent marketplaces, machine-to-machine payments, and user-owned AI assistants. The idea is to combine blockchain-native accounts, payments, and asset ownership with the ability of software agents to act automatically.
Compared with crypto projects that moved toward AI only after the sector became fashionable, NEAR has a clearer technical link to the theme. Co-founder Illia Polosukhin previously worked on machine learning research at Google and was involved in the landmark Transformer paper, Attention Is All You Need. The other co-founder, Alexander Skidanov, has a background in distributed databases and large-scale systems.
That gives the AI narrative more continuity than a simple marketing relabeling. According to project disclosures, NEAR’s AI ecosystem includes more than 50 teams across research, data, storage, models, and applications. Partnerships or integrations mentioned in the article include Frax, Infinex, SWEAT, and Eliza. NEAR is also advancing AI assistants such as IronClaw, with a focus on trusted execution environments, credential isolation, and privacy protections so agents can safely access accounts, hold assets, and execute transactions.
There is also a natural overlap between AI agents and Intents. In the future, software agents may need to buy services, manage capital, and exchange assets across multiple blockchains on their own. Intents could help them find routes and liquidity, while NEAR could become the coordination and settlement layer for those machine-driven transactions.
The addressable vision is large. The business evidence is still thin. More than 50 AI teams does not mean more than 50 mature revenue streams, and partnership announcements do not automatically mean recurring paid demand. Centralized AI services already have established compute, tools, and enterprise customers. NEAR still has to show that blockchain-based ownership, privacy, and verifiability can outweigh the extra technical complexity.
More importantly, even if AI products succeed, the resulting value may not accrue directly to the NEAR token. Agents may access services through third-party wallets, application teams may cover fees on behalf of users, and end users may never need to buy or hold NEAR over time. Two questions remain unresolved: whether AI activity can produce protocol revenue, and whether that revenue can create token demand.
$13 billion in settlement is not the same as protocol income
The central issue in the current NEAR thesis is not transaction scale on its own. It is the gap between transaction scale and protocol revenue.
The source article cites a revenue panel showing that NEAR-related products generated about $3.48 million in total fees in the 30 days before Sept. 1, 2026. After payments to solvers, partners, applications, and other participants, net protocol fees were about $757,500.
In other words, the protocol keeps only a portion of the total fee pool. That split is not surprising in cross-chain transactions. Solvers need to quote prices, market makers need to provide liquidity, and wallets and applications need to earn revenue too. But for token holders, gross settlement and total fees are not the most important numbers. Net protocol revenue is closer to the economic value the underlying asset may actually capture.
Even then, net fees should not be treated as the equivalent of shareholder cash flow. NEAR can direct revenue toward buybacks, staking, lockups, ecosystem spending, or the protocol treasury, and there is no fixed rule requiring all revenue to be used to purchase tokens. Capital allocation remains subject to governance choices.
That means user counts, transaction counts, or Intents settlement figures can overstate the quality of growth if viewed in isolation. The article argues that three questions matter more:
- Is net protocol revenue growing on a sustained basis?
- Is the share of total fees retained by the protocol improving?
- How much of that retained revenue is converted into buying, burning, or long-term locking demand for NEAR?
If Intents settlement rises quickly but net revenue stays relatively low, NEAR could become widely used infrastructure without becoming an asset that captures value efficiently.
Token economics have improved, but the loop is still incomplete
NEAR has taken steps to address that issue. Its maximum annual inflation rate has been reduced from 5% to about 2.5%. The developer gas rebate mechanism has been removed, allowing a larger share of transaction fees to move through the burn process. NEAR Intents has also activated a fee switch and has used part of product revenue for buybacks. According to project disclosures, these mechanisms have already led to buybacks of more than 1 million NEAR.
The supply picture has improved as well. About 1.305 billion NEAR tokens are already fully or nearly fully in circulation, leaving the circulating market capitalization and the fully diluted valuation broadly aligned. Compared with newer chains that still face large token unlocks for teams and investors, NEAR carries less concentrated unlock pressure ahead.
Still, being close to fully circulating does not mean the asset is deflationary. NEAR fees remain low, so even with high transaction counts the amount burned may not be enough to offset issuance to validators. Buybacks also involve discretion and have not yet become a stable, predictable, rules-based framework.
So the standard for proof is higher than showing that token economics are better than in the previous cycle. NEAR needs to show that protocol revenue and fee burns can eventually approach or exceed new issuance. Without that, product growth does not automatically flow through to token value.
User and developer metrics point to activity, not yet category leadership
NEAR’s ecosystem figures look strong on the surface. In its 2024 review, the project said monthly active users rose from 7 million in the prior year to 40 million, daily active users were about 4 million, and average daily transactions exceeded 8 million. Ecosystem projects raised about $146 million in outside funding that year, and accelerator participants raised another roughly $50.5 million.
At a minimum, those figures show that NEAR is not a ghost chain that has lost its developers and users. After the last bear market, it still retained a network, infrastructure, capital, and an application ecosystem.
But blockchain user metrics need careful interpretation. Monthly active users may refer to accounts, addresses, or application interaction entities, and those numbers can include automation, subsidy-driven activity, and low-value transactions. They should not be read as equivalent to unique users on a traditional internet platform.
The article also notes that it did not obtain complete, independently verified current data on retention, stablecoin scale, total value locked, or transaction quality. So 40 million monthly active users and 8 million daily transactions show activity. They do not, by themselves, prove that the activity produces recurring revenue.
The same pattern appears in developer data. The article cites August 2026 figures showing about 1,231 developers, 79,400 code commits, and 234 repositories in the NEAR ecosystem. On the same basis, Solana had about 1,494 developers while Ethereum had 11,600.
That suggests NEAR still has a functioning developer base, but it remains well behind Ethereum and also trails Solana in liquidity and consumer application momentum. Applications such as HOT Wallet, SWEAT, and KAIKAI have helped build a user base on NEAR, yet none has become the kind of breakout product that changes the industry’s competitive map.
Competition stretches across several markets at once
NEAR is not competing in just one arena. In base-layer blockchains, it is up against Ethereum, Solana, Sui, and Aptos for developers and capital. In chain abstraction, it faces wallets, trading aggregators, bridges, and other intents networks. In AI, it must also contend with centralized cloud providers and crypto projects focused on compute, data, models, and privacy.
That mix gives NEAR both meaningful optionality and a high execution burden. If its product lines reinforce each other, the project could build a broader system around accounts, payments, cross-chain transactions, and AI agents. If resources become too scattered, it could end up with many products and only limited market power.
At this stage, the article frames NEAR as a technically credible mid-sized contender with a differentiated direction rather than a market leader that already has a deep moat.
The real test is no longer technology alone
The most compelling version of the NEAR investment case is that it becomes coordination, transaction, and settlement infrastructure for a multi-chain economy and for AI agents. If that happens, the market NEAR can address may be much larger than that of a single Layer 1, and the project could build new paths to value capture through Intents fees, agent payments, and token buybacks.
Its strengths are clear enough: an experienced founding team, a long operating history, a sharded architecture, a token structure that is close to fully circulating, lower inflation, and a differentiated mix of AI and chain abstraction.
Its largest risk, however, may not be technical failure. It may be remaining stuck in a state where the technology works, the product set keeps expanding, but the economic position stays secondary. NEAR could continue to operate smoothly, maintain a developer community, and launch new products without ever building a strong enough application network, protocol revenue base, or token demand engine.
The article says the next metrics worth watching are NEAR Intents net revenue and margins, retention among non-incentivized users, stablecoin and TVL growth, real commercial revenue from AI products, effective token inflation, and transparency around buybacks and treasury funds.
In the end, three questions define the debate. How much revenue can Intents transaction growth actually leave with the protocol? Will AI and cross-chain products create direct demand for NEAR tokens? Can protocol revenue, burns, and buybacks offset new issuance over time?
Until those questions are answered with stronger evidence, NEAR appears better suited to an optionality trade on AI and multi-chain infrastructure than to a core asset backed by established cash flow and a proven network moat. The upside case rests on its technology and product direction. Whether that turns into a durable rerating still depends on commercialization, value capture, and competitive pressure.