TL;DR
NEAR in 2026 does not look like a chain that died in one dramatic blow. It looks like a project bleeding relevance slowly. The core layer-1 case lost force: economic weight is modest, the old growth narrative faded, and the market increasingly treats NEAR as peripheral rather than central. The one serious counterargument is its AI and chain-abstraction stack. That is the part of the story that still looks strategically alive. So the real question is not whether NEAR vanished. It is whether the AI pivot is a reinvention or simply the last respectable explanation for why the market should still care.
Published March 18, 2026. Updated March 18, 2026.
Disclosure: This page is editorial analysis based on publicly available protocol materials, infrastructure updates, market data, and third-party research. A consolidated source list appears in Sources & Notes near the end.
Jump to:
- Is NEAR dying slowly?
- The short answer
- What changed since the old bull case
- Why the decline looks gradual, not terminal
- Core-chain economics vs. AI narrative
- Why the AI pivot is the only serious counterargument
- What would change the verdict
- FAQ
- Sources & Notes
The cleanest way to describe NEAR in 2026 is not “dead,” and it is not “thriving” either. It looks more like the blockchain equivalent of an old internet brand that still exists, still has infrastructure, still has a user story, but no longer feels like where the future is being decided.
That distinction matters because sudden failure and slow irrelevance are different diagnoses. A chain can survive for a long time after it stops feeling strategically central. That is the harder argument here: NEAR did not suffer a quick kill. It suffered a slow bleed.
The original layer-1 pitch was strong on paper: fast finality, sharding, lower friction, better usability, and a founder set with real technical credibility. But markets do not reward good architecture automatically. They reward ecosystems that become gravitational. NEAR has not obviously done that. What it has done instead is pivot hard toward AI, chain abstraction, and intents. That may be a reinvention. It may also be the last credible explanation for why the market should still keep it on the shortlist.
Is NEAR Dying Slowly? The Short Answer
Yes, that is the better reading. NEAR in 2026 looks less like a chain that collapsed and more like one that gradually lost strategic relevance.
The bearish case is not that nothing works. Parts of NEAR do work. The bearish case is that the market stopped treating the original NEAR thesis as a top-tier destination. The core chain remains small relative to major competitors, the ecosystem no longer feels culturally central, and the strongest current activity increasingly sits in the chain-abstraction and intents layer rather than in the old “this L1 wins on fundamentals” story.
So the real 2026 verdict is narrower and more useful: NEAR looks weaker as a standalone L1 winner than it once did, but still has one plausible survival path through AI-native infrastructure and cross-chain execution.
What Changed Since the Old Bull Case?
What changed is not just price. It is how the product is being explained.
An earlier bullish NEAR article could center the chain itself: scalability, accessibility, ecosystem growth, and the idea that superior architecture would eventually convert into dominant adoption. In 2026, that framing looks incomplete. The protocol’s own messaging is now much more explicit about a different future. NEAR’s chain-abstraction page says the goal is to eliminate blockchain complexity so AI can interact with assets and applications across chains “as if they were a single system” NEAR chain abstraction. Its intents stack is framed as an AI-native transaction layer for moving value across Web2, Web3, and traditional markets NEAR Intents.
That is not a small positioning tweak. It is a strategic tell. When a chain increasingly sells the abstraction layer instead of the base-layer victory story, it usually means the old pitch did not become inevitable.
The operating backdrop changed too. In January 2024, the NEAR Foundation said it would cut roughly 40% of staff to focus on a narrower and higher-impact set of activities The Block on NEAR Foundation staff cuts. In May 2025, NEAR announced the phased deprecation of free public RPC endpoints under `near.org` and `pagoda.co`, explicitly noting that it followed Pagoda winding down operations and the ecosystem moving toward a more sustainable infrastructure model NEAR RPC deprecation notice.
None of that proves collapse. It does show contraction, refocusing, and a chain that no longer looks like it is expanding from unambiguous strength.
Why the Decline Looks Gradual, Not Terminal
The reason “slow bleed” is the right frame is that NEAR still has enough life to avoid a clean obituary. It still has infrastructure. It still has institutional memory. It still has technical differentiation. It still has some measurable activity. That is exactly what makes the Yahoo/AOL comparison useful: the issue is not immediate disappearance. The issue is relevance decay.
Markets usually make this kind of judgment quietly. First a project stops feeling like the obvious next winner. Then attention moves elsewhere. Then the story becomes conditional: “interesting if the pivot works,” “worth watching if adoption returns,” “still technically strong, but…” By the time everyone says the category has faded, the drift happened long before the obituary.
We have written before about how Web3 often confuses surface motion with durable positioning, whether through manufactured traction signals or more general optics-first operating behavior. NEAR’s problem in 2026 is less theatrical than that. It is more structural. The chain no longer feels like a default destination for capital, builders, or mindshare.
That matters because crypto does not just reward technical merit. It rewards strategic gravity. The winners become where liquidity settles, where developers concentrate, where adjacent infrastructure compounds, and where outside observers assume the next wave will happen. NEAR increasingly looks like a chain people can explain, but no longer instinctively prioritize.
Core-Chain Economics vs. the AI Narrative
This is where the case gets uncomfortable. The strongest evidence that NEAR has been bleeding relevance is not rhetorical. It is economic.
DefiLlama currently shows the NEAR chain at roughly $92.47 million in DeFi TVL, with just $2,139 in 24-hour chain fees and the same amount in 24-hour chain revenue DefiLlama NEAR chain page. Even allowing for the limits of TVL and fee metrics, that is not what strategic dominance looks like. It is what a peripheral chain looks like.
The token side tells a similar story. DefiLlama’s protocol page for NEAR shows a market cap of roughly $1.35 billion against an all-time high price of $20.44, with the token still far below the level where the old market imagination once placed it DefiLlama NEAR protocol page. Price alone is not destiny, but it is often a blunt market verdict on how much strategic belief has survived.
And yet there is a twist. The most interesting current metrics do not sit in the core chain story. They sit in NEAR Intents. DefiLlama shows NEAR Intents with roughly $58.65 million in TVL, about $3.74 million in fees over the last 30 days, and about $1.817 billion in 30-day DEX volume DefiLlama NEAR Intents. That is not proof that NEAR has won. It is proof that the one part of the story still generating real strategic interest is not the old monolithic L1 thesis.
This is the Ben Thompson version of the argument: the market is effectively telling NEAR where it may still matter. It is not rewarding NEAR for being a cleaner layer-1 in the abstract. It is paying more attention when NEAR acts like infrastructure that simplifies cross-chain complexity for agents, applications, and users.
The AI Pivot Is the Only Serious Counterargument
If you want the bullish case in 2026, it has to run through AI, chain abstraction, and intents. Anything else feels stale.
NEAR’s own materials make that clear. The protocol says chain abstraction lets AI interact with assets and services across multiple chains as if they were one system, and that NEAR Intents is designed so users or AI agents can express outcomes while the runtime handles routing and settlement NEAR chain abstraction and NEAR Intents. In plain English: NEAR is no longer just trying to be a better chain. It is trying to be the coordination layer that hides the chain map altogether.
That is strategically smarter than pretending the market will simply re-run the old L1 competition. It also gives NEAR a cleaner answer to a real 2026 question: what does blockchain infrastructure look like if AI agents need to transact across fragmented systems without making users think about bridges, wallets, and rails?
But this is also where the skepticism has to stay sharp. An AI pivot can be reinvention. It can also be a respectable new wrapper around an old relevance problem. Plenty of crypto projects now want to borrow AI’s momentum. The bar is therefore higher, not lower. NEAR does not just need an AI narrative. It needs evidence that the AI-native layer becomes economically meaningful in a way the old base-layer story never fully did.
That is why this page does not dismiss the pivot, but it also does not grade it on branding. In Web3, that mistake is common enough that we built broader frameworks around how real verification should work and what stronger standards should actually test. The same rule applies here: interesting architecture is not the same thing as durable market proof.
What Would Change the Verdict?
If NEAR wants to escape the “slow bleed” framing, it has to prove more than technical competence. It has to show compounding strategic relevance.
That would look like a few concrete things:
- Core economic improvement: materially stronger fees, revenue, and retained activity at the chain level, not just cleaner messaging.
- AI-native product pull: evidence that agents, apps, or services are choosing NEAR’s abstraction layer because it is operationally better, not because the narrative is fashionable.
- Cross-chain defensibility: proof that intents and chain abstraction create switching costs or compounding data/network effects rather than acting as interchangeable middleware.
- Clearer operating maturity: less ecosystem contraction language, more repeatable evidence of durable infrastructure, governance, and business traction.
Until then, the default reading stays the same: NEAR is still here, but the old winner’s aura is gone. The remaining question is whether the AI layer becomes a genuine second life or simply a more sophisticated way of slowing the fade.
FAQ: NEAR AI Blockchain Review 2026
Is NEAR dead in 2026?
No. The better description is that NEAR looks strategically weaker and more peripheral than it once did. It still has infrastructure and a live product story, but the decline looks gradual rather than explosive.
Why call NEAR a slow bleed instead of a collapse?
Because NEAR still functions. It still has technical differentiation and ongoing development. What changed is relevance: the market no longer treats the original layer-1 thesis as obviously central, and the strongest current story sits in AI and chain abstraction instead.
What is the strongest bullish argument for NEAR now?
The strongest bullish case is that NEAR’s chain-abstraction and intents stack becomes useful infrastructure for AI agents and cross-chain execution. That is the one part of the story that still looks strategically fresh.
What is the main bearish argument against NEAR?
That the core chain has modest economic weight relative to bigger competitors, the old growth narrative lost credibility, and the AI pivot may be a reinvention attempt rather than proof the original thesis won.
Is NEAR’s AI pivot real or just marketing?
It is real enough to take seriously, because the protocol has built around chain abstraction and intents and current metrics show meaningful activity there. But it is not yet strong enough to erase the broader relevance problem.
Sources & Notes
- NEAR chain abstraction — accessed March 18, 2026.
- NEAR Intents — accessed March 18, 2026.
- Deprecation of NEAR.org and Pagoda.co RPC Endpoints — accessed March 18, 2026.
- The Block: Near Foundation to cut staff by 40% — published January 11, 2024; accessed March 18, 2026.
- DefiLlama NEAR chain page — accessed March 18, 2026.
- DefiLlama NEAR protocol page — accessed March 18, 2026.
- DefiLlama NEAR Intents — accessed March 18, 2026.
Disclaimer
This report is for general information and editorial analysis only. It does not constitute legal, investment, tax, or business advice. Digital-asset risks and metrics change quickly; readers should verify current facts directly with primary and official sources.
What Would Actually Work: The Real Conditions for NEAR AI Pivot
The most honest question to ask about any startup pivoting to a new narrative is: does the pivot solve the original problem, or does it replace the original problem with a different story? NEAR pivot to AI is the most interesting strategic bet in the L1 space right now — not because it is obviously correct, but because it is testable. There are specific conditions under which it works. Most of the current analysis does not specify what those conditions are.
The original problem NEAR was trying to solve was developer experience on blockchains. The thesis was that Ethereum was too hard to build on, and that NEAR sharding architecture and JavaScript-friendly SDK would attract the next generation of developers who wanted blockchain functionality without the pain of Solidity. That thesis was partly right and mostly too early. The developer experience problem was real. NEAR solution was genuine. But the market was not large enough, and Ethereum ecosystem had enough gravitational pull that most developers who needed blockchain functionality chose to build on Ethereum or Ethereum-compatible chains rather than migrate to a new environment.
The AI pivot is a different thesis. The claim is not that NEAR is easier to build on. The claim is that NEAR architecture is specifically suited to running AI agents and AI-adjacent applications on-chain — that the combination of scalable execution and a user-owned data model creates something the AI industry needs but cannot get from cloud infrastructure. That is a substantive claim and it deserves a substantive evaluation.
Here is where the open-source AI competition changes the calculus. DeepSeek and Qwen releasing capable models with permissive licenses removes one of the primary blockers for deploying AI on blockchain — cost. Two years ago, running meaningful AI inference on-chain was prohibitively expensive because the models required significant compute. Today, small, efficient models derived from the open-source wave can run inference at a cost that makes on-chain AI applications economically viable. NEAR did not create that condition, but it is a genuine tailwind for the pivot thesis.
The enterprise question is the bottleneck that the pivot thesis does not yet answer. Enterprise AI adoption is happening primarily through managed cloud services — Microsoft Copilot, Google Vertex, Amazon Bedrock — where the enterprise gets AI capability without owning the infrastructure or the model weights. The pitch for on-chain AI would need to solve a specific enterprise problem that managed cloud cannot solve: user data ownership at scale, auditability requirements, or interoperability between AI systems that are currently siloed behind different cloud providers. NEAR has articulated this pitch. It has not yet demonstrated it at a scale that enterprise buyers would find compelling.
The comparison to Berachain Proof-of-Liquidity approach is instructive for a different reason. Berachain solved a specific bootstrapping problem — how to get liquidity on a new chain without relying purely on speculative token incentives — by tying block rewards to on-chain economic activity. NEAR AI pivot would benefit from a similar mechanism: something that ties NEAR token economics to actual AI agent usage rather than speculation about future usage. The whitepaper for NEAR AI roadmap describes user-owned data as the value capture mechanism, but the specific pathway from data ownership to NEAR token demand is not yet clearly defined.
The market is making bets on this question in real time. Prediction markets on AI blockchain adoption have not yet developed enough depth to provide reliable signals on NEAR specifically, but the aggregate market probability on AI-native blockchain applications reaching meaningful enterprise adoption by 2027 is below 20%. That is a market judgment, not a verdict — markets are often wrong on long timelines, especially in technology transitions. But it tells you what the aggregate sophisticated bettor thinks about the base rate.
The VC signal is clearer. Crypto venture capital in 2025-2026 has concentrated on infrastructure plays with clearer near-term revenue pathways — Hyperliquid, on-chain credit protocols, stablecoin infrastructure — rather than on AI blockchain thesis plays. NEAR has raised capital and has institutional supporters. But the frontier VC is not betting heavily on the AI blockchain category as a 2025-2026 investment. That changes if NEAR can show a live enterprise deployment with verifiable usage metrics, not just developer interest at conferences.
What would actually change the verdict: a Fortune 500 company publishing a case study on running AI agents on NEAR in production, with specific data on cost savings or capability advantages over managed cloud alternatives. That is the evidence that would shift the probability from “interesting experiment” to “real business.” Until that evidence exists, the AI pivot is a thesis with genuine intellectual coherence and insufficient validation. Both things can be true simultaneously — and the distinction matters for how you size a position.
The Civilizational Frame: What Blockchain AI Protocols Need to Understand About How Technologies Actually Win
Yuval Noah Harari’s framework for understanding how technologies achieve civilizational adoption is built on a single observation that the technology industry consistently underweights: technologies do not win because they are technically superior — they win because they become the shared fiction that coordinates large groups of humans toward common purposes. Money is not valuable because of its physical properties; it is valuable because a sufficient number of humans agree to treat it as valuable. Nations are not powerful because of their geographic boundaries; they are powerful because their citizens share a coherent enough collective identity to act in coordination. The blockchain and AI protocols that will achieve genuine civilizational adoption are not the ones with the best technical architecture — they are the ones that succeed in becoming the shared coordination layer that humans choose to organize around, for reasons that are partly technical and partly narrative and partly historical accident.
Harari’s analysis of NEAR AI’s challenge is that it is competing in a context where the shared fictions have already partially consolidated. Ethereum is not technically superior to every challenger — but it is the blockchain that the largest number of institutions, developers, and protocols have chosen to build around, and that collective choice is self-reinforcing in ways that technical improvements alone cannot overcome. The developer who builds on NEAR is making a bet that NEAR’s technical advantages — the sharding architecture, the AI integration layer, the user experience improvements — will attract enough coordination that the network eventually becomes a genuine alternative to the Ethereum ecosystem’s collective weight. This is a bet on narrative and coordination, not just technology. And Harari’s historical record of which technologies win suggests that the bet requires the challenger to offer not just a better technical solution but a compelling alternative shared fiction about what the technology is for and who it is for.
The AI pivot that NEAR is attempting is the most interesting strategic question in the civilizational frame: if AI and blockchain are both becoming civilizationally significant technologies, is there a shared fiction available that combines them in a way that creates a new coordination layer rather than a derivative of the existing ones? The candidate shared fiction is something like “verifiable AI intelligence operating on an open, transparent, user-owned infrastructure” — a story that responds to the specific civilizational anxiety about centralized AI control that Harari’s own work has helped create. If the anxiety about who controls AI is real and growing, then the protocol that most convincingly embodies the alternative to centralized AI control has a narrative advantage that technical architecture alone cannot create. Enterprise AI’s centralization concerns — the vendor lock-in risk, the data sovereignty question, the alignment of the AI model with the enterprise’s interests rather than the model provider’s — are the specific anxiety that a blockchain-AI protocol like NEAR is positioned to address if the narrative is constructed correctly.
Harari’s historical perspective on why technologies fail to achieve the adoption their technical merits would predict identifies the credibility gap as the primary failure mode: the technology may work, but the shared fiction about what it does and who validates it has not been built. The blockchain industry has a specific credibility gap problem that is structurally different from the AI industry’s credibility gap: AI has been adopted at civilizational scale in ways that blockchain has not, and the reason is partly that AI’s benefits were immediate and observable while blockchain’s benefits required the adoption of a new coordination layer before the benefits could be experienced. Crypto’s press release problem is the failure mode of the credibility gap: the industry has defaulted to promotional assertion rather than the demonstration-and-validation cycle that builds genuine shared fiction. Independent editorial credibility is the mechanism through which a blockchain AI protocol builds the kind of verified shared fiction that Harari’s framework identifies as necessary for civilizational adoption — Wikipedia notability is the digital-age proxy for the institutional legitimacy that transforms a technology from a niche coordination tool into a shared infrastructure. Developer platform lock-in is the specific civilizational risk that NEAR’s open infrastructure narrative needs to address directly: the developers who have been squeezed by Microsoft’s developer platform extraction are the audience for whom the open blockchain AI infrastructure story has the most immediate resonance. Prediction markets on NEAR’s developer activity growth through end-2026 are pricing modest improvement rather than the inflection that civilizational adoption would require — which Harari’s framework reads as the market’s honest assessment that the shared fiction has not yet been built convincingly enough to create the coordination that would make the technical advantages self-reinforcing.
