Idle Computing Power, Cheap Electricity, and Overlooked Models — DGrid Aims to Turn Them into Revenue
If you have a batch of idle GPUs or are in an area with extremely low electricity prices, can you make money with AI? What if you fine-tuned a model that performs exceptionally well in a specific vertical but have no channel to sell it?
These seemingly unrelated questions have the same answer provided by DGrid's newly launched Model Marketplace: connect them to an open model market, allowing every bit of computing power, every kilowatt of electricity, and every good model to find users willing to pay for them.
The number of AI models has exploded at an astonishing rate over the past two years. From general large models to fine-tuned versions for vertical industries, from code generation to multimodal understanding, the supply side has become extremely rich. However, as the number of models increases, the challenge of "getting models used and ensuring that those who create models and provide computing power actually earn money" has not become easier.
For those capable of creating good models or holding computing resources, monetization remains a poorly addressed issue; for developers, finding "the right model for this scenario" still means jumping between various platforms. The Model Marketplace officially launched by DGrid targets this middle ground where both supply and demand sides are complaining but have yet to find a good solution.
An Open Market, Not Just Another API Transfer Station
To understand the positioning of the DGrid Model Marketplace, one must first distinguish between two different business logics.
Most current "AI model aggregation" platforms operate on a middleman logic: the platform itself connects channels, negotiates discounts, and then repackages and sells to users. This model has an inherent limitation — users and developers are completely unaware of what the platform can access, how it prices, and who the suppliers are; it is essentially a black box.
DGrid takes a different path: an open market. Any model provider — whether a team creating original large models, an organization fine-tuning for vertical scenarios, or a deployment party with computing infrastructure capabilities — can directly list models on the platform, set their own prices, and earn on-chain direct settlements when the models are called.
The platform itself does not survive by taking a cut but acts as infrastructure to facilitate supply and demand. Conceptually, this is closer to what a healthy market should look like than "black box transfers." Its true imagination lies in the answer to the question of "who can become a supplier."
Eight Ways to Turn Resources into Revenue
The DGrid Model Marketplace is open not just to model developers. It aims to incorporate several previously segmented or long-idle resources into the same market for monetization.
Path for Idle Computing Power. For organizations or individuals with GPU infrastructure but underutilization, deploying models and providing services externally through DGrid is a direct monetization path. Computing power is no longer just a cost but can be transformed into a sustainable revenue-generating asset.
Internal Computing Power Overflowing Externally. Many enterprises or teams deploy models for internal use, but the actual usage often cannot fill all the computing power. DGrid provides a mechanism: while meeting internal needs, it opens up the remaining computing power and service capabilities to external users, turning "loss-making fixed costs" into partially variable revenue.
Distribution Channels for Vertically Fine-Tuned Models. Models fine-tuned for specific scenarios such as law, healthcare, finance, and code often outperform general large models on specific tasks but struggle to reach users who truly need them. DGrid provides standardized listing and distribution paths — model teams only need to focus on making the models good, while distribution and billing are handled by the platform.
Lightweight Distribution for Large Model Developers. For large model teams with R&D capabilities, distribution and market operations are often non-core but resource-consuming tasks. By listing on DGrid, they can directly reach a large number of paying developers and users, focusing on the model's capabilities without having to repeatedly invest in channel building.
Price Advantages for Channel Resource Providers. Some suppliers hold specific model channel resources or have lower cost structures, allowing them to offer more competitive prices on DGrid than official sources, attracting users and forming differentiated competition. This kind of robust market competition means lower usage costs for users.
Better Model Discovery Mechanism. For developers, the current dilemma is not just "too many models to choose from," but also that "good models are buried in information noise and cannot be found." DGrid's unified market provides standardized model descriptions, capability tags, and quality signals verified by PoQ, making it easier for high-quality models to be discovered and recommended.
Cross-Regional Calling and Compliance Flexibility. The availability of AI models is often affected by regional restrictions — certain models cannot be accessed directly in specific regions, or some users have multi-regional calling needs. By calling through DGrid's unified entry, cross-regional access issues can be resolved within a compliance framework, broadening the actual availability of models.
Arbitrage from Electricity to Computing Power Across Regions. This may be the most imaginative scenario: in areas with extremely low electricity prices (such as some regions rich in hydropower), deploying AI inference services connected to DGrid essentially transforms low-cost electricity into AI computing power, which is then output to global users in the form of model calling services. This provides a new path for "cross-regional electricity arbitrage" and allows global users to enjoy AI services at a lower cost.
These eight scenarios cover a complete supply-side map from computing power providers, model developers, channel resource providers to regional arbitrageurs. The value of the DGrid Model Marketplace lies in aggregating these previously dispersed, inefficient, or idle resources into a clearly defined, real-paying market.
PoQ: The Trust Issue That Open Markets Must Solve
Openness brings rich supply but also introduces a new problem: when anyone can list models, how do users know that the models they are calling are genuine?
This is precisely the purpose of DGrid's PoQ (Proof of Quality) mechanism.
A key detail needs to be clarified: the object of PoQ verification is the model provider, not each call made by users. Its working method is: DGrid independently and randomly samples the providers on the platform using its own benchmark test set, then publishes the verification results on-chain for public access. The entire process does not touch user request data and does not put user data on-chain.
This design addresses the "substandard goods" problem that is most likely to arise in open markets. Providers know they may be sampled at any time, and the cost of quietly downgrading models is detection and punishment, rather than just relying on self-discipline. PoQ provides a trust foundation for this open market from a mechanistic perspective.
The core members of the DGrid team have doctoral backgrounds from institutions such as Stony Brook University and have published four academic papers on PoQ on arXiv, covering core designs of Proof of Quality, Optimistic TEE-Rollups, Cost-Aware Proof, and PoQ-Judge, among others. This is uncommon among similar platforms.
In contrast, current high-market-cap AI aggregation platforms (such as OpenRouter) have brand endorsements but lack any on-chain verification mechanisms. DGrid's PoQ replaces "trust-based" with "verification-based," solving the problem at a mechanistic level.
The Demand Side Has Proven, Now the Supply Side is Next
The launch of the DGrid Model Marketplace is not about creating a brand new market but about unlocking the supply side of an existing market with real demand.
From the demand side, DGrid has already aggregated over 200 mainstream models, including Claude, GPT, Gemini, MiniMax, and GLM, with more than 15,000 paying users and cumulative revenue exceeding $23 million in the first half of 2026 — all achieved based on $5 million in seed round financing. This set of numbers indicates one thing: a large number of users are continuously paying for AI services, the demand is real, and the market exists.
However, most of the existing 200+ models are top commercial models that DGrid actively integrates. The real gap on the supply side lies in those with genuine capabilities, vertical models, fine-tuned models, and infrastructure parties with computing resources but lacking monetization channels; previously, there was no standardized channel to enter this market, nor a clear revenue-sharing mechanism. The Model Marketplace was built precisely for this gap.
DGrid has validated the willingness to pay on the demand side with $23 million in revenue, and now it is expanding the supply side with an open market, allowing both supply and demand to match on the same platform, and using PoQ to ensure quality — the logic of the flywheel is clear.
Several Key Points Worth Observing Next
The success or failure of any open market ultimately depends on execution. For DGrid, there are several directions worth continuous attention — these are both its opportunities and the key to testing whether this model can work.
First, the speed of supply scale expansion. The value of an open market lies in the aggregation effect: the more models there are, the richer it becomes, making it more attractive to developers; the more users there are, the more returns for suppliers. DGrid's existing 15,000+ paying users and real calling demand give it a clear advantage over platforms starting from scratch — once suppliers settle in, they face an existing market that is already paying. The number of high-quality vertical models and computing resources it can attract next will determine how quickly this flywheel turns.
Second, the ongoing effectiveness of the quality screening mechanism. The more open it is, the greater the quality disparity among incoming models, which is a common issue for all open markets. DGrid's differentiation lies in that it does not rely solely on manual review but has the PoQ on-chain verification mechanism as a fundamental filter. As the scale expands, whether PoQ can consistently distinguish between high-quality and low-quality models will be the core of this market's long-term credibility.
Third, the integration of AI and Crypto users. DGrid connects both Web3 native users and a large number of traditional AI developers who only care about "how good the model is and whether it is cheap." Serving both groups well is a common challenge in bridging AI and Crypto products, and DGrid's product design of "on-chain settlement but seamless experience" has already brought 15,000+ paying users onto the same platform — this itself is a good signal.
Conclusion
The landscape of the AI model market is still forming. Leading commercial models will continue to iterate, but there is a vast space of vertical scenarios, specific use cases, and a large amount of undervalued computing resources that have yet to be fully tapped.
The significance of the DGrid Model Marketplace lies not only in whether it can immediately become the "Taobao of AI models" but in the broader question it raises: if idle computing power can be transformed into AI services, low-cost electricity can become globally available inference capabilities, and fine-tuned models can directly reach the users who need them most — how large could this market be?
DGrid chooses to answer this question with an open market and an on-chain verification mechanism. For model teams looking for distribution channels, infrastructure parties holding computing resources but lacking monetization paths, and observers concerned about the progress of the DeAI track, this release is worth a serious look.
Disclaimer: This content is provided for general branding and informational purposes only and doesn't constitute financial, investment, legal, or tax advice. Any events, rewards, online events, or related information mentioned herein should not be considered a recommendation, solicitation, or invitation to purchase, sell, trade, or otherwise deal in any crypto assets or to use any services. Crypto assets are highly volatile and may result in loss. WEEX services and online events may not be available in all regions and are subject to applicable laws, regulations, and eligibility requirements. You are responsible for ensuring that your use of WEEX services complies with local laws and for carefully assessing the risks before participating in any crypto-related activities.
You may also like

Soitec's Stock Soars 24%: Why This French AI Gem Is Shaking Up the Stock Market

Tokenized Stocks on the Rise, Yet Traditional Market Infrastructure Faces Severe Challenges

Decoding SK Hynix's Cross-Market Arbitrage Wave: A Brick-Moving Frenzy Spanning US Stocks, Korean Stocks, and the Crypto Market

Pendle Unveils H2 Roadmap: Focus on RWA and Institutional Markets, Boros to Emphasize On-Chain Interest Rate Derivatives

Why Korean Companies Choose Hong Kong for RWA Expansion

5 leading Bitcoin-backed loan platforms in 2026

Ethereum Reserves Decline on Exchanges, but Demand Lags Behind

Crypto ETF: Ethereum Surpasses Bitcoin as BlackRock Takes the Lead

In-Depth Analysis and Defense Guide for Fastjson 1.2.83 'Gadget-Free' Vulnerability (0day)

Ethereum vs BNB Whitepaper Comparison (2026)

Founder of Primitive Ventures: In the Age of AI, Those Who 'Disappear' Are the Underclass

Revolving Door Trading Exposed: Who is Tailoring the U.S. Stablecoin Bill for Tether?

The New York Times: Founders at Odds? The Battle Between Kalshi and Polymarket is More Intense Than Expected

The Sandbox and Animoca Brands Host $10,000 AI Competition in Hong Kong

Agentic Payment from Visa's Perspective

European Central Bank Expected to 'Pause Hawkish' Tonight, Door for Autumn Rate Hike Remains Open

Mainnet Approaches: A Comprehensive Overview of Circle's Native Blockchain Arc Ecosystem

From Issuance to Revenue: Uncovering the Hidden Gold Mine in the Stablecoin Trillion-Dollar Market

Next AI Investment Target: Cryptocurrency, Franklin Templeton Suggests

Retail Dividend Fades, Predicting an AI Arms Race in the Market

Swiss Bank BancaStato Launches Bitcoin Trading Through Sygnum And Avaloq

Bitcoin: Long-Term Holding Reaches Historic High

Workers at Ukraine's Largest Chemical Plant Threaten Strike Over Three-Year Salary Arrears

IBM Lowers Full-Year Revenue Guidance, Can It Still Price Based on Stable Cash Flow?

BTC Returns to $66,000: Is This a Sign of Recovery?

Demand for Auto Insurance Increased by 100-200% Amid Ukrainian Shelling

Dual Throat Crisis Approaches: Markets Face Energy Shock and Long-Debt Pressure

Glassnode Conducts Investigation Amid Data Leak Concerns, Warns Customers of Phishing Risks

AI Agents Drive Up Server CPU Space, Funds Pulling Out of Tech Stocks

