Original Author: Eric, Foresight News
Recently, Solana made a joke about Starknet, mocking an L2 with only 8 daily active users and 10 transactions per day that still has a $15 billion FDV.
Although, in hindsight, this salt-in-the-wound joke was meant to attract attention and introduce the news of Starknet's token STRK launching on Solana via NEAR Intents. But Solana's criticism is not without reason; the L2s that have sprung up like mushrooms in the past two years are indeed facing a traffic crisis.
The most convincing recent example is on January 8th, when Zero Network, an L2 network incubated by Web3 wallet company Zerion, was exposed to have stopped producing blocks for over 3 weeks, but it seemed to have no impact. The official reaction was even more subtle; Zero Network stopped producing blocks on December 19, 2025, but the official statement on fixing the issue only came on December 23. The last original content posted by Zero Network's official Twitter was in May of the same year.
Despite this, the claim of only 8 users executing 10 transactions per day is an exaggeration. According to data from L2BEAT, Starknet's TPS yesterday was 2.64, meaning there were over 200,000 transactions on the network in a day. But this number is still extremely low; even the Ethereum mainnet's daily transaction volume is 10 times that of Starknet.
Data shows that among general-purpose L2s, aside from Base and Polygon, even Arbitrum and OP Mainnet's TPS are not significantly higher than Ethereum's. Linea and Starknet have a TPS of less than 3. The parts not shown in the screenshot include Scroll with a TPS slightly over 1, and ZKsync, Blast, etc., with less than 1.
Looking at the TVL data from DefiLlama, Base and Arbitrum together account for nearly 80% of all L2 TVL. The remaining L2s, not in the "Others" category, had a combined valuation conservatively estimated at nearly $10 billion during their private funding rounds, but their combined TVL is less than $2 billion.
In terms of protocol revenue, only the top 7 protocols had revenues above $1,000 in the past 24 hours. Daily protocol revenues in the three-digit or even two-digit range might not even compare to the interest some large holders earn daily from placing assets in exchange wealth management.
This data very直观地展示了 (intuitively shows) the current predicament of L2s: against the backdrop of a lack of application narratives, hoping for a killer app that doesn't operate as an appchain and willingly runs on a general-purpose L2 has become a pipe dream. On the question of finding an application scenario that can provide stable transaction data, L2s have found the same answer: cryptocurrency cards.
Pavel Paramonov, founder of cryptocurrency research institution Hazeflow, once criticized that cryptocurrency cards are essentially not "cryptocurrency payments" but still fiat payments, not truly promoting cryptocurrency adoption. But he also mentioned that many projects or public chains launching crypto cards are doing so out of necessity, with the goal merely being to keep users within their ecosystem.
Many cryptocurrency cards launched by exchanges currently are a form of "custodial" card. User assets are held in exchange or institutional custody accounts, and upon consumption, settlement is handled by the exchange, off-ramp companies, and the card issuer. The settlement chains for such cards are usually Tron, Solana, or even the slightly higher-cost Ethereum. On one hand, the存量 (存量 - inventory/supply) of stablecoin assets on these chains is sufficient; on the other hand, some cards reduce costs through batch settlement rather than per-transaction settlement. For institutions, liquidity and stability might be more important than the low cost of L2s.
The crypto cards that L2s are eyeing are various forms of "non-custodial cards." Before using such a card for payment, assets reside in the user's own wallet, and each payment is settled individually, which can effectively increase on-chain activity. Typical examples include Scroll (Etherfi card settlement layer), Gnosis, and Linea (MetaMask card settlement layer).
In September 2024, Etherfi announced that its payment card would use Scroll as the settlement layer. Scroll can help Etherfi achieve "gasless transactions" and provide a higher cashback percentage subsidized by the SCR token. Besides the traditional method of directly spending assets on Scroll, the Etherfi card has a special mechanism: users can borrow fiat currency using yield-bearing assets on Scroll as collateral for payments. Supported assets include eETH, weETH, wETH, eBTC, etc.
Gnosis, as a sidechain that has long lacked presence, has successfully made a comeback with its payment card. Its card, Gnosis Pay, primarily operates in Europe. Users can connect non-custodial wallets like MetaMask and Gnosis Safe in the Gnosis Pay App. During spending, Gnosis Pay converts the user's wallet's supported assets (certain Euro, Pound, and Dollar stablecoins) into EURe, a Euro stablecoin issued by Monerium, which is then converted 1:1 into Euros for payment.
The crypto card issued by MetaMask uses ConsenSys's L2 Linea as the primary settlement network, additionally supporting Solana and Base. Users need to deposit supported payment assets (various USD or Euro stablecoins) into their MetaMask wallet before spending. During payment, user assets are transferred to the off-ramp service provider, converted into fiat currency, and then paid to the merchant.
Due to the per-transaction settlement nature of non-custodial cards, each user消费 (消费 - consumption/spending) corresponds to a contract call to verify the remaining asset balance and the on-chain asset transfer. Thus, L2s can rely on payments, an absolutely high-frequency and sustainable scenario, to ensure a certain level of on-chain activity. According to Paymentscan data, Scroll, through its partnership with Etherfi and SCR subsidies, has captured a significant market share in card payments. However, this data is not entirely accurate, as many card payments might not involve on-chain transfers but are settled internally by institutions. Regardless, it's an indisputable fact that L2s have found a practical application scenario through payments.
It's not just emerging L2s that are anxious; Polygon, which cannot be strictly classified as an L2, also recently shifted its strategic focus to payments. By the end of 2025, the transfer volume of non-USD stablecoins on Polygon exceeded $11.1 billion, with the new stablecoin XSGD reaching $2.24 billion in trading volume and the Australian Dollar stablecoin AUDF reaching $2.46 billion. Additionally, Polygon has become one of the main chains used for Stripe's stablecoin payments; its announcement on January 13th to acquire cryptocurrency payment infrastructure Coinme and blockchain development platform Sequence for a consideration of $250 million更是将 (更是将 - further writes) "all in on payments" plainly on its face.
After experiencing the bombardment of various concepts, L2s have come to recognize reality. While they still await novel applications, the urgent task at hand is to utilize their low-cost, high-efficiency characteristics and survive first through payments.
