NEAR: Ultrasound Markets
$NEAR is pursuing Ethereum’s programmable-money ambition across markets it doesn’t need to own or build. An idea far bigger than another blockchain competing for developers.
It asks the question...
What if your entire financial life could run through one account, with AI connecting you to the world’s markets?
Successful exchanges, applications and chains become places it can connect to, with each integration bringing new assets, services, and sources of revenue. A market of markets, with AI and a seamless UX as the interface.
Sounds powerful.
Disclosure: I hold NEAR. Figures refer to the September 2026 research snapshots cited below.
1. Ideology
NEAR believes that in a world that evolves towards maximum efficiency, the economy will trend towards AI at the frontend, with markets and blockchains powering the backend. You express what you want to do; AI coordinates the execution, with hyper-efficient and interconnected blockchains providing the infrastructure through which money moves.
In this world, NEAR believes that privacy is fundamental. Consult an AI about your holdings, shift your daughter’s savings account into a new asset, or simply pay for a coffee, and without privacy you leak financial information that can be used against you. Users seek convenience, but they don’t want to give every service and every interested observer a copy of their financial life.
Whereas ZEC enabled privacy for a store of value, NEAR is attempting to extend confidentiality into the exchanges, services and computation involved in putting that money to work.
2. Origin Story & Team
NEAR has one of the most OG origin stories in crypto.
While at Google in 2017, @ilblackdragon co-authored Attention Is All You Need, the paper that introduced the Transformer architecture behind modern LLMs. Architecture that quite literally powers the AI systems you use today such as ChatGPT, Grok, Claude and Gemini.
That same year, believing that a major use case for AI would be software development, Illia joined forces with @AlexSkidanov , an engineer who had worked on distributed databases at MemSQL. With a background in machine learning and scalable systems, they combined expertise and founded NEAR AI to direct AI’s capabilities towards software development.
As often is the case with good ideas, they found themselves at the right place, but at the wrong time. The models and available computing resources were not yet ready for the ambition. Paying contributors internationally also exposed another problem: access to banking and the difficulty of moving money across borders.
With AI powered software development parked aside, the founders turned their attention to payments with the goal of building a blockchain with scalability designed in from the outset. Over the past few years, NEAR has evolved from a more scalable L1 into an ecosystem tackling some of the premier problems users encounter onchain: gas, cross-chain complexity and publicly broadcast transactions.
3. Technology
NEAR’s current product suite can be explained through three key elements:
- NEAR(dot)com: A platform bringing crypto, tokenised assets, earning and trading into one connected experience.
- NEAR Intents: Cross-chain execution infrastructure connecting onchain markets and assets.
- NEAR AI: Confidential, verifiable AI infrastructure for people, businesses and their agents.
Near.com
NEAR.com is a single interface for managing money across tokenised markets.
You can create a wallet with a passkey, swap between supported chains, hold tokenised stocks and gold, send payments and trade perpetuals through one interface, with gas and bridging abstracted away underneath.
For someone like me, that means being able to move part of an onchain portfolio into other assets without coordinating several wallets, bridges and exchanges. For someone new, it means creating a wallet with Touch ID and getting access to markets in minutes. Its confidential account also enables you to keep balances and activity private, while letting you share records when needed.
NEAR Intents
Intents is what allows this experience to extend beyond NEAR.com.
An application requests an outcome, such as exchanging BTC for ZEC, and market makers compete to fulfil it. This architecture also allows developers to request quotes, execute swaps and track completion through an API, rather than needing to choose, build or manage the underlying liquidity infrastructure.
I tested this while building tipz.cash, a private tipping application. I wanted creators to receive tips in shielded ZEC without requiring every supporter to first acquire it. Intents let me build around the experience I wanted: the supporter pays with a supported asset they already hold, and the creator receives the asset they want. I didn't need to think about liquidity or the best venue for trading, I just plugged into an interface that was connected to them all.
Confidential Intents, wrap a confidential layer around this architecture offering restricted visibility for users. They currently account for 6% of lifetime volume, but that share is rapidly growing with over 23.7% over the past month and 29.2% over the past week. They also held $113.25 million, slightly more than half of total Intents TVL.
NEAR AI
NEAR AI provides confidential computing infrastructure for applications that need to work with sensitive information. @AskVenice uses its private inference infrastructure, showing that the customer does not need to be managing money, or even know they are using NEAR, for the technology to be useful.
An agent managing your finances needs access to sensitive information and a way to act on it. NEAR AI provides private inference and hosting, while Intents provides access to supported markets.
Bringing those capabilities together could make an assistant useful across more of your financial life: understanding your holdings, preparing a rebalance or coordinating a payment within the permissions you have given it. NEAR.com could provide that experience through one account, while other applications use the same infrastructure to serve their own customers.
4. Competitors
NEAR faces competition at each part of its product suite: the account, the execution infrastructure and private AI.
Centralised exchanges already bring balances, trading and transfers into a convenient account, with a user base of millions. But… their custodial services require handing control of your assets to an operator that can also see your activity.
NEAR.com’s proposition is to bring that convenience onchain, with user control and confidential balances and transactions. Its planned AI tools would extend the account from somewhere you execute decisions into somewhere you can understand and manage your finances.
Protocols like CoW and Thorchain compete on solver and cross chain markets, but lack the full product suite and interface that Near provides.
It's evident that NEAR’s potential advantage is connecting all its capabilities into a single interface for financial execution: a venue where users, agents, and developers can seamlessly plug in to access the world of tokenized finance. Or... "a market of markets".
5. Tokenomics
NEAR is powered by a proof of stake consensus mechanism, with 90% of issuance allocated to validators/stakers and 10% to the treasury.
The current supply sits around 1.307 billion, with no fixed maximum. An issuance of 2.5% implies roughly 32.7 million new tokens a year before burns.
With around 42.92% of supply staked, user can receive a staking rate of approximately 5.24% a year before validator fees. Users can also direct staking rewards towards private inference through NEAR AI.
70% of smart-contract execution gas fees are burned An approved upgrade would raise that to 100%, though the latest cited update still had mainnet activation pending. The overall burn dynamic purely from a blockchain perspective is questionable; NEAR was built to make transactions cheap. That helps users, but it takes considerable activity for gas burns to absorb new issuance.
Intents provides a source of revenue with more potential. Users can move between assets on different chains without ever choosing to interact with NEAR directly, while the fees retained from their transactions fund NEAR purchases in the background.
The fees generated via intents varies with the source, but a portion of all is utilised to buy back NEAR. The September 21 dashboard snapshot reported $1.89 million retained from $5.39 million in fees over 30 days, which would execute to roughly $23 million annualised if that pace continued.
This is where the “ultrasound money” idea becomes relevant. In Ethereum’s version, the concept was the network fee burns could exceed new issuance.
NEAR could pursue a similar outcome without being limited to its's or othe Layer 2 networks with competing governance dynamics. With fees intertwined with intents, Near generates revenue from every chain and every app it plugs into.
Intents buybacks are held rather than burned, so the underlying ultrasound mechanism potentially requires a tokenomics evolution, but the potential is there.
6. Macro Relevance
Tokenisation is still small relative to the markets it aims to bring onchain. A September 15 snapshot put distributed tokenised assets at approximately $38.8 billion, excluding stablecoins, according to Eco’s reporting of RWA.xyz data.
Forecasts see this growing rapidly. McKinsey estimates $2 trillion by 2030, Citi’s puts it at $5.5 trillion, and BCG projects $14 trillion.
For me, the interesting question is how we access all of it. I want to move between stocks, funds and gold without learning a new system every time I move my money. Standardisation is one of tokenisation’s biggest advantages, but putting assets onchain doesn’t automatically connect their liquidity or make them available through the same account.
At the same time, we are seeing people use AI to help decide where their money goes. Adobe found that 39% of surveyed US consumers had used AI for shopping, while traffic from AI sources to US retailers grew 393% year over year in early 2026. That is still largely assistance rather than autonomous spending, but it gives us an early indication of how the interface is changing.
McKinsey estimates that agents could mediate $3–5 trillion in annual consumer goods sales by 2030, before including services or business-to-business commerce.
7. Risks and Headwinds
The risk is that NEAR correctly anticipates where technology, AI and commerce is going, while others capture most of the economics.
Exchanges, brokers and payment platforms can add tokenised assets and AI tools directly to their existing products, giving users more capabilities without giving them a reason to leave. A much larger onchain economy could develop while most customers continue accessing it through familiar providers.
Established platforms already connect merchants, consumers and payments, and can choose which infrastructure sits underneath. NEAR could become a supplier to those businesses, but winning distribution will likely require revenue share agreements.
Centralised exchanges could also remain the simplest option for managing assets. Self-custody and confidentiality appeal to me, but many users may prioritise familiar interfaces, account recovery and customer support. NEAR needs to make its advantages easy enough to use that they influence where people actually keep and transact with their money.
8. The Investment Case
NEAR has built a product I want to use. The question is whether it can make that infrastructure equally useful to the applications and agents other people use.
It seems obvious that AI will expand into managing financials. The challenge is making NEAR one of the easiest places for an agent to transact. Connecting an API is the beginning; developers also need reliable execution, clear permissions and confidence that financial information and credentials remain protected. Illia has explicitly identified agent security as a prerequisite for this economy to develop.
Distribution is also key. Wallet integrations put Intents in front of existing crypto users, but the real expansion is into products whose customers dont even know they are using a blockchain, let alone Near.
There is an early example in Abound, a financial application for Indians abroad. NEAR AI announced a pilot for agents that monitor accounts and prepare or execute remittances under user-defined conditions. Abound reported more than 800,000 users and $300 million in processed remittances, illustrating a unique angle of attack that near can provide; Provide infrastructure to a business that already has customers and a financial problem to solve.
Commerce could follow the same pattern. A checkout, a business purchasing services or an agent paying for data may need to exchange what the payer holds for what the recipient accepts. NEAR’s Agent Market already explores transactions beyond financial trading, although meaningful commercial adoption remains to be demonstrated. Banking and card features could also extend the usefulness of the consumer account alongside these integrations.
This gives the investment case a sequence we can actually assess: a useful product, easier agent access, wider distribution, then recurring revenue.
NEAR looks small beside established financial platforms, but that alone doesn’t make it cheap. Coinbase and ICE generate billions in revenue from trading, services and financial infrastructure. Their value comes from monetising customer relationships across several products.
For NEAR, the distinction is whether it remains a replaceable swap provider or becomes infrastructure people and businesses repeatedly depend on. NEAR.com and private AI could broaden what customers pay for, but those revenues only support the investment where value actually reaches the token.
Conclusion
I started with a question: is there an asymmetric investment hiding underneath a product I want to use?
NEAR is pursuing an open financial system by connecting markets, with AI potentially making them accessible to a much wider audience. The product makes that ambition tangible, but its ultimately down to the integrations and revenue economics that will determine how much that intersection is worth.
Easier agent access and wider distribution could bring the customers; retained revenue and token purchases have to complete the investment case.
Connecting trillion-dollar markets through a single interface, powered by private AI built by the 'co-founder of LLMs;, and sustained by a potential revenue flywheel from the markets, networks and products it integrates…
How would you value that?
"Ultrasound markets"
Get free research here: https://bloomresearch.substack.com/