Grass, part 2: the next big meta in DePIN + AI?

Big companies are forecasting a 10x growth in the AI market. If you believe AI is the next big thing, $GRASS stands out as the go-to crypto play. Here’s why:

I. What is Grass?

1. Overall

Grass is a decentralized platform that enables users to monetize their unused internet bandwidth. This bandwidth is utilized to scrape and structure web data for AI training purposes, with users receiving rewards and fee-sharing in return.

Setting up a Grass node is straightforward and doesn’t require powerful hardware. Users can operate Grass nodes on laptops (downloading app) or Android, integrated with Solana Saga phone (currently testing). By keeping their devices connected to the internet and running, users can earn rewards. Grass ensures that it doesn’t affect users’ connections, as it only utilizes unused bandwidth and scrapes public data without accessing personal information.

2. Tech

Grass is running a Layer 2 Rollup, based on Solana as Layer 1.

The model looks like this (picture below). Overall, whenever Grass nodes scrape data from the web, validators track the exact data origin and the responsible node, storing all this information on-chain. This ensures that users are rewarded proportionately while preserving the integrity of the data scraping process, helpful for AI training later.

→ In short: Users use Grass via Nodes (super easily installed) and get rewarded proportionately via Grass network, a Solana L2.

3. Team

Grass is developed by Wynd Labs. Currently, the team remains mostly anonymous, with only Co-Founder Andrej R and CTO Chris Nguyen publicly identified. Some minor issues related to information disclosure that need attention.

4. Funding

• Jan 2023: Raised $1 million in pre-seed funding from No Limit Holdings and Big Brain Holdings.
• Dec 2023: Raised $3.5 million in seed round led by Polychain Capital and Tribe Capital, with participation from Bitscale Capital and Big Brain Holdings.
• Sep 2024: Completed Series A funding round led by Hack VC, with contributions from Polychain Capital, Delphi Digital, Lattice, etc. While the exact amount was not disclosed, sources indicate that valuation was at nearly $1 billion.

Total: $4.5M

5. Tokenomics

• Ticker: GRASS
• Max supply: 1,000,000,000
• Network: Solana

II. Why Grass could be the next big thing

1. Good Traction + Big Revenue Expectation

a) Good Traction

In just one year since launch, Grass claims to have hit 2.5M DAUs, and that number has now grown past 3 million. This isn’t just empty hype — on-chain data confirms 2.1M users claimed the airdrop, backing the claim to some extent.

According to Grass Foundation, there have been 109M IP addresses connected from 190 countries. If we take the 3M DAUs at face value, that’s about 2.7% of IPs being real active users — which sounds pretty reasonable.

For referrence:

• Arbitrum currently has ~1.7M active addresses
• BNB Chain has ~1.3M

So if these numbers hold, Grass is already one of the most used chains/dapp in the entire space. Not to mention the revenue.

b) Big Revenue Expectation

Grass’s business model revolves around scraping web data and selling it to AI companies for model training. So the more data they scrape, the bigger the potential revenue.

Looking at the chart, before Jan 2025, daily data scraped averaged ~10TB. But after the launch of the Sion update (an infra upgrade that made scraping multimodal data far more efficient — up to 10x), daily data spiked massively, hitting an ATH of 1,700TB and now averaging ~1,000TB daily.

Scraping volume is directly tied to user count. Typically, projects see user drop-offs post-airdrop since users usually pay to farm. But in Grass’s case, users earn passively without cost, so users could actually retain even after airdrop.

Let’s break down the forecast based on conservative assumptions:

• User base stays stable (no major increase or drop)
• In Q1 2025, Grass scraped 57,000 TB. With the current rate of 1,000 TB/day, each quarter could hit ~90,000 TB, so 57k is conservative. Annualized: 57k x 4 = 228,000 TB = 228M GB.
• Assume only 10% of that data is useful and can be sold.
• Market is estimating $1 — $5/ GB for data to train AI model, Photobucket offered $1/ video; Adobe asked users to submit videos for $2.62 each. So we take a conservative $1/ GB.

→ Annual revenue = 228M GB x 10% x $1 = $22.8M

Even under conservative assumptions, $GRASS revenue is already on par with top crypto protocols like Pendle, Lido, and Arbitrum. In a recent demo, the founder also casually mentioned a “mid-8 figure revenue” — let just say $50M annually — which seems realistic considering the conservative case already estimates ~$22M.

This puts $GRASS not only in the top tier of crypto projects, but also in the same league as big global AI companies. Yet it’s still only sitting at a $400M market cap, while AI companies with similar revenue get valued at $1B–$5B.

However, to be real and not just hopium, some fundamental questions still need answers:

• Who’s buying this data?
• Is the $1/GB pricing verified and consistent?
• Is it one-off or recurring revenue?

That said, in crypto, narratives and expectations often front-run fundamentals. And this story is strong enough to pump on narrative alone, especially with the AI angle.

2. Attractive AI narrative

Common sense: Compared to other AI projects, Grass’s product is really easy to understand and has a huge addressable market, because anyone with internet and a device can be Grass users. Furthermore, AI data industry is a $100B industry, without question about its potential.

Big expectations:

• Grass claims that it has been reached out by Fortune 500 companies.
• Co-founder claims that Grass can scrape the whole internet on a daily basis, which is now only possible with 2 companies, Microsoft and Google.
• Rumor of partnerships with Xbox and Roku TV.

Most of these are rumors and cannot be verified. However, these rumors are the ‘Big if true’ type, and really easily conceivable (relate to web2 giants), which is good in my view. As crypto is driven by speculation, speculation is driven by expectation and awareness, which are all good in this case.

Also, when comparing $GRASS to other top AI coins with similar fundamentals, VC backing, and awareness, it clearly looks undervalued. AI gem trading at a 50–70% discount — a pretty juicy deal.

3. Favorable tokenomics

Obvious that this cycle has a hate toward cabal tokenomics, and favor over tokenomics that looks “good for the community”.

Grass has:

• 10% for community airdrop, and promising 17% for futures airdrop, which is a large amount.
• 50% for the team & investors with 1-year cliff. Locked tokens cannot be staked, avoiding dumping staking rewards like TIA, SUI.

While the circulating supply at TGE is 25.86%. The real initial supply is even better:

• As only 70M airdrop token claimed, 30M left.
• 32.6M tokens ~ 42% or $GRASS have been staked.

And for the coming months, main emissions will be:

• 6M staking reward
• 2.71M monthly in Ecosystem fund

Meaning 3.1M monthly emission, with current price of $1.5 → max $4.5M dump monthly. With current average volume on Bybit ranging $5M -$15M daily, even the worst case won’t affect price much.

4. LCR — Product USP

LCR (Live Context Retrieval) is Grass’s USP compare to other AI project, a core engine that supplies real-time internet data to other AI models. It aggregates data scraped from all Grass nodes, allowing other systems to plug in and access this live data — a potential game-changer.

The Pain Point? Yes. In an experiment, a base AI (with outdated data) and a search-augmented AI (requiring live data) were asked 100 questions on recent events. The base AI struggled, trying to infer context from conversation, often leading to incorrect answers. In contrast, the search-augmented AI answered 98/100 correctly by simply looking up the information. A recent paper on ChatGPT also showed that, in 40% of queries, only 1% of the available data was used, claiming the out-dated data useless.

The Demand? Absolutely. As a common sense, we can see clearly that all current applications need real-time data to solve problems. Fields like finance, healthcare, and social marketing, etc would face serious issues without updated data. ChatGPT generates $1.2B in annual revenue through ChatGPT Plus, offering up-to-date data and advanced models. According to a KX report, real-time data services could add $4.2 trillion globally. As AI continues to grow, so does the need for constant data. If there were an index for global real-time data demand, the chart would be up-only.

The Supply? Scarce. Although most companies need real-time data, few can afford the infrastructure to build it due to high costs. While giants like Google, Microsoft, and Amazon have this capability, they surely won’t share data with potential competitors. If LCR succeeds, it could become the “missing piece” that fulfills all companies’ needs. As being said, if Grass could scrape the whole internet daily like claimed, make it the 1 and only real-time data open source. In short, in the most bullish case Grass can be the PMF for the whole AI world.

Summary:

Product-wise, Grass is a high-quality project that has found product-market fit, with strong traction across both web2 and web3 users, putting it in the top tier of dapps. It’s backed by expectations of massive revenue potential and rides the AI mega-narrative.

Right now, it’s still a story play — but one of the only crypto products (besides WLD) with a big enough story to back it. It taps into real web2 demand, in a trillion-dollar industry with up-only growth and no clear leader yet.

On top of that, it’s currently undervalued by 30–60% compared to top AI coins, and has no major unlocks in the next 6 months.

Conclusion: $GRASS is a solid narrative + fundamental-backed play, and worth the investment.

Research conducted by Monolith team.

Subscribe to our socials for more alpha: X, LinkedIn.

Website: https://monolith.vc/