Preface

Solana has been around for five years, with Jito (the leader in MEV infrastructure on Solana) developing in less than three years. However, its market share has rapidly grown from the initial 15% to the current 95%. It can be said that most of the Meme buy-and-sell transactions on Solana have to go through it! No advertisements here, so please sit back, hit the favorite button, and let’s dive deep into the underlying principles as I gradually reveal:

1. What is Jito-Solana? Why has it captured the market in two years?
2. How does Solana’s core mechanism differ from Ethereum’s?
3. Why is your transaction often getting squeezed?
4. How will the MEV landscape on Solana evolve in the future?

1. What is Jito-Solana?

Of course, Jito-Solana is not the only provider of MEV infrastructure on Solana (with a 95% market share). There are others like Paladin, Deeznode, BlockRazor, BloxRoute, Galaxy, Nozomi, and so on, each with different entry points. This article will focus on the development process and technical principles of the core leader, and later discuss the advantages, disadvantages, and entry points of these other companies.

1.1. The Development Timeline of Jito

Let’s take a look at the rapid market share development through a timeline, focusing on the staking rate and related partners.

• Established at the end of 2021
• Launched on Solana mainnet in June 2022, with 200 validators by September, covering 15% of the staking volume
• 2022-2023: Financing, iteration, and cooperation with Solana Foundation, with Jito client being included in the official recommendations
• 2023 TGE: Staking Jito to earn MEV rewards, forming a restaking model
• Q1 2024: Due to strong community opposition, Jito-Solana closed the channel to send transactions to Jito-Blockengine
• Q2 2024: Over 500 validator collaborations, covering 70% of Solana’s MEV, with 3 billion transactions processed in 2024
• Q1 2025: Staking coverage reaches 94.71%. Today, the importance of cross-chain bridges remains undeniable.

[Source: https://www.jito.network/zh/stats/]

Therefore, it can be said that Jito is the leading infrastructure provider in the MEV ecosystem on Solana. In the past three years of development, it has built a solid foundation of support from Solana validators, making it so that the vast majority of transactions must go through Jito’s system.

It is through Jito’s system that Solana’s downtime has significantly decreased.

It is Jito that has enabled sandwich attackers to earn high profits.

It is also Jito that has provided Solana validators with an additional 30% MEV rewards, consistently and steadily.

Moreover, Jito has transformed from the initial dragon-slayer into the dragon itself, shifting back and forth between being the hero and the villain, sometimes fierce, sometimes benevolent.

In today’s mainstream Meme narrative, Jito has become a two-faced operator, excelling in both directions.

1.2. What Exactly Has Jito Built?

It is actually composed of three core services: block-engine, jito-solana, and jito-relayer. Their relationships are shown in the diagram below:

[Source: Created by Shisi-Jun]

First is block-engine, which is an auction system. Its typical scenario allows sandwich attackers to submit a bundle of up to five transactions in a fixed order for auction. Additionally, they can send an extra transaction, called a Tip, to validators as a small incentive to prioritize packaging this Bundle. Other scenarios include decentralized exchanges (DEX) platforms like OKX, GMGN, and BN Wallet. To avoid users being targeted by sandwich attacks, they can add a tip to individual users’ transactions, following the auction route to get their transactions processed faster.

Next is jito-solana, a client that replaces validators for transaction validation and block creation. Its core function is to allow validators to receive Bundle packages sent by block-engine, enabling them to prioritize processing these transactions and completing the transaction sequence. At its peak, the system processed up to 25 million Bundles daily (currently around 10 million), with each bundle typically being profitable. The tip fees are ultimately collected by a unified account and then distributed to validators (95-97%) and Jito itself (3-5%).

The most controversial component is jito-relayer, which can be understood as a gateway for validators to receive transactions. Initially, when this relayer received a transaction, it would delay by 200ms before forwarding it to jito-solana, while simultaneously forwarding it to block-engine with no delay. Clearly, this was selling order data. Thus, the rise of Jito in its early days was due to the profit gap caused by user losses. It’s important to note that in March 2024, the official statement said they no longer transmit data, but as of today, the jito-relayer still contains the switch and setting for the 200ms delay. So, whether validators are selling user data today remains unclear, especially since the block-engine remains closed-source.

1.3. How Much Has Jito’s System Earned?

Clearly, the profit gap between transactions being targeted or not, combined with the introduction of the tips mechanism, has ultimately benefited validators, contributing to Jito’s rapid market share growth. Who would turn down an additional 30% in earnings?

In the past year, a total of 4.3 billion Bundles were initiated, and the total Tips fees amounted to 5.51 million SOL. At a market price of 140 USD per SOL, this resulted in an additional revenue of 7.7 billion USD for Jito’s infrastructure.

[Source: https://explorer.jito.wtf/]

However, not all of these revenues go to Jito the system provider. As mentioned earlier, Jito shares 3-5% of the platform fees with validators, so Jito’s actual earnings over the past year amounted to around 200,000-270,000 SOL, roughly 35 million USD.

This is comparable to the revenue of King of Glory in two days, which might not seem high at first glance. However, this is a solid platform profit, especially considering that most Web3 industries cannot even point to specific revenues.

Jito has become a monopolist, with exclusivity over other competitors (since validators can only run one client), and its earnings have recently been impacted by meme fatigue. If Solana can explore more transaction scenarios in the long term,

and if there are no market competitors, the platform may adjust its revenue sharing from 3-5% to 30% (which is a common platform fee rate for internet applications once they dominate the market).

This could lead to a very high PE estimate. Using a 30 PE ratio common for Web2 industry leaders, the valuation could reach 1 billion USD. Using the more typical 300x multiple for Web3 industry monopolies and potential growth, the valuation could soar to 10 billion USD. Similar valuation methods are referenced in “Super Intermediary or Business Genius? Revisiting the Year After LayerZero’s Transition from V1 to V2.“

However, today’s focus isn’t on such macro conclusions, nor on understanding Jito through speculative virtual valuations. Instead, we aim to delve into the details, understand the deeper principles behind it, and analyze the future market development.

1.4. What Demand Scenarios Can Jito Support?

This topic essentially revolves around the most common types of MEV attacks, the most frequent of which is Frontrunning, such as:

• Arbitrage: Like Ethereum, it’s a no-risk arbitrage operation.
• Sandwich Attack: A typical sandwich attack, where the profit on Solana for a sandwich trade is about 2 USD per transaction.
• JIT (Just in Time) Liquidity: Providing operations that offer instant liquidity.

Another major category is Backrunning: Involves inserting an arbitrage trade after a target transaction (such as a large DEX trade or liquidation event) to profit from the market volatility caused by that target transaction. Specific scenarios include:

• DEX Arbitrage: You can think of this as taking advantage of price differences between different DEXs. The trade would follow immediately to close the arbitrage gap.
• Liquidation Follow-up: After a user’s collateral is liquidated, the follow-up trade buys the asset at a discounted price and resells it.
• Oracle Delay: Executes reverse operations based on outdated prices before the oracle updates its data.

Apart from obvious attack scenarios, there are other acceleration scenarios that also benefit from Jito. Therefore, it’s fair to say that Jito isn’t only for MEV, but also serves all scenarios with bundling needs for accelerated or batch transactions.

For example, during the busy token launch activities on Solana, traders often use the bundling and acceleration mechanisms to open markets and deploy tokens.

Similarly, large exchanges could also use the bundling of tips for large user transactions to avoid attacks. However, it’s important to note that these measures can’t prevent validators from malicious actions (and in fact, you can’t be sure which validator is acting maliciously).

2. In-Depth Comparison of Solana and ETH System Differences

Why is Jito so well-suited for Solana? Why doesn’t this market have the same level of competition as ETH, with multiple contenders? To understand this, we must look at the system differences between the two. You may have heard of the POH consensus many times, but Solana’s transaction lifecycle is different from Ethereum’s, and this creates a stark contrast between their ecosystems.

2.1. ETH’s MEV Landscape

Two years ago, on the first anniversary of Ethereum’s Merge, I conducted a systematic analysis titled “MEV Landscape One Year After Ethereum’s Merge.” In this analysis, the clear difference in Ethereum’s system lifecycle was evident:

[Image Source: https://mp.weixin.qq.com/s/IepFvVpIxLpkXV5qgF68Rw]

This is due to two important changes after the Merge:

1. Ethereum’s block interval became stable. It was no longer the previous random interval between 3 and 30 seconds. This stability had both positive and negative effects on MEV. On one hand, searchers didn’t have to rush to send slightly profitable transactions but could wait to accumulate a better transaction sequence before submitting it to validators. On the other hand, it intensified competition among searchers.

2. Miner incentives were reduced. This encouraged validators to accept MEV transaction auctions more willingly, leading to MEV reaching a 90% market share in just 2-3 months.

As a result, we have roles like Searcher, Builder, Relay, Proposer, and Validator in Ethereum’s MEV ecosystem.

The lifecycle of each block is as follows:

1. The builder creates a block by receiving transactions from users, searchers, or other (private or public) order flows.
2. The builder submits the block to the relay (where there can be multiple builders).
3. The relay verifies the validity of the block and calculates how much it should pay the block proposer.
4. The relay sends the transaction sequence and reward price (auction bid) to the current block proposer.
5. The proposer evaluates all received bids and chooses the highest reward sequence.
6. The proposer sends the signed block header back to the relay, completing the auction process.
7. After the block is published, the rewards are distributed to the builder and proposer based on the block’s transactions and the block reward.

Thus, I believe that Ethereum’s MEV situation inevitably places searchers and builders in fierce internal competition.

Actual data supports this view: the overall yield has decreased significantly by 62%.

• The average profit before the Merge (from September 2021 to September 2022) was 22MU/M, including arbitrage and liquidation modes.
• The average profit after the Merge (from December 2022 to September 2023) was 8.3MU/M, including arbitrage and sandwich modes.

You might argue that MEV-boost on ETH has increased faster. Indeed, it has!

However, faster growth doesn’t necessarily mean higher profits. As we’ve already analyzed, Ethereum’s growth is due to reduced miner incentives, which encouraged validators to accept MEV transaction auctions, enabling MEV to capture 90% of the market share in just 2-3 months.

[Image source: https://mevboost.pics/]

The key difference between ETH and Solana lies in the presence of multiple builders, each with different final revenues that influence validator decisions. This creates competition among builders.

Due to this competition, the profits of searchers are constantly compressed. Beyond algorithms, the only differentiator for builders becomes the amount of data they can access.

Searchers who can’t compete will exit the market, while builders with access to large volumes of data often have their own infrastructure and market reputation, forming a stable flow of total orders rather than relying on Mempool propagation between nodes.

As a result, compared to Solana’s more centralized and oligopolistic platform model, ETH’s MEV market is more market-driven, offering users relatively more breathing room.

2.2 Solana’s Block Mechanism

After understanding the ETH system, please clear your mind—many aspects of Solana are different from Ethereum, even down to the fundamental concept of blocks.

These very mechanisms are at the root of the rampant MEV activity in the Solana system.

We can use a table to quickly compare these four core features.

The key lies in two features: the absence of a mempool and direct connections to the leader. The former introduces transaction latency, while the latter opens the door to validator misbehavior.

2.2.1 Solana Actually Has No Mempool

Previously, we mentioned Jito’s 200ms delay and its syncing with the block engine. That’s essentially a monetized mempool. But in principle, Solana doesn’t use a mempool—this is part of its design for speed and privacy. So how does this affect the block production process?

If you’re a regular user sending a transaction to a node, that’s effectively a broadcast. Under default settings, the node will immediately locate the current and next leader (two validators in total) and forward your transaction. But where is the transaction order determined? That depends on whether you’re using native Solana or Jito-Solana:

• Native Solana: Once the leader receives the transaction, it theoretically follows a FIFO (first in, first out) model. Whichever transaction reaches the leader first gets included in the sequence. Combined with the POH mechanism, this divides a block into many small ticks for synchronized processing.
• Jito-Solana: Transactions go into a queue that calculates the current gas limit (called CU, or Compute Units, in the SVM system). This queue has lower priority than Bundled transactions. So regular transactions end up after Bundles. If someone attacks you with the same transaction, Jito-Solana will prioritize bundling the attacking transactions. Here, 80% of block space is reserved for Bundles, and only 20% goes to regular transactions.

So, Solana doesn’t eliminate transaction propagation—it just reduces public propagation. This makes Solana’s MEV landscape more exclusive, favoring high-end players.

2.2.2 Future Leaders Are Predictable

Validators are selected per epoch (roughly every 2–3 days) from the pool of 1,300 validators, using a VDF-based random sampling mechanism that’s weighted by stake.

For example, if 2 million SOL is staked in total and you’ve staked 200,000 SOL, you have a 10% chance of being selected. If selected, you’ll produce blocks for the next 4 slots (Solana’s equivalent of blocks), which lasts about 1.6 seconds.

Because this process is predictable and fast, any active node can calculate who the upcoming validators will be and attempt to connect with them in advance to submit user transactions. Due to network latency, it’s common for transactions to miss the current leader and arrive at the next.

2.2.3 Leader Connection Strategy Is Also Stake-Weighted

This is handled through the SWQoS (Stake-Weighted Quality of Service) mechanism. A leader has a total P2P connection capacity of 2,500 peers. Of these:

80% (2,000 connections) are reserved for nodes that have staked SOL (SWQoS participants).

20% (500 connections) are available for non-staked public nodes.

It may sound complicated, but it’s essentially an anti-spam and Sybil-resistance mechanism designed to prioritize transaction messages routed through staked validator proxies.

2.3 Why Is It Easy to Get Attacked on Solana?

Many regular users wonder: can I avoid getting sandwich-attacked by offering a high Priority Fee to get my transaction packed first by validators? The truth is—it helps a bit, but not much. In extreme cases, it may even backfire.

[Image source: https://explorer.jito.wtf/feestats]

As the chart shows, Priority Fees offer a probabilistic advantage, while tips are more volatile and competitive. Also, tips are separate transactions—externally, it’s unclear which ones are in a Bundle.

So even with a high Priority Fee, your transaction only moves up in the validator’s remaining 20% queue for that slot. But if a Searcher spots your order early and launches a sandwich attack using a Bundle (which includes your transaction), your high Priority Fee simply makes that Bundle more valuable (higher average CU cost), increasing its chance of being processed early within the validator’s Bundle queue and broadcasted quickly.

Similarly, other Solana mechanisms might seem user-friendly at first glance—but Solana still has one of the most aggressive MEV environments. Here’s why:

• Leaders Can Misbehave Without Accountability

Leader slots are consecutive, meaning leaders A and B can both access all user transactions. This lowers the cost and ambiguity of misbehavior for leader B.

Imagine you’re leader B and spot a profitable user transaction. You can quickly construct a sandwich attack, send it to the block engine for auction, and under the 80% Bundle-priority rule, your attack likely gets processed first—even though it’s packed by leader A.

How would anyone prove that leader B is the attacker?

You might argue leader A should be blamed for including the attack, but in practice, 95% of validators would behave the same, so there’s little incentive or mechanism for punishment.

• Transaction Retries and Long Delays

Each slot is only 400ms, yet many users report transactions stuck for over 23 seconds. Why?

It’s not poor node performance—it’s SWQoS. If you’re connected to a regular (non-staked) node, it may find the correct leader but fail to connect during congestion. With only 500 connection slots reserved for such nodes, your transaction may fail and enter a retry loop every 2 seconds.

These are default parameters (some nodes tweak retry intervals), but as of March 2025, there are about 1,300 validators and 4,000 RPC nodes. During congestion, 2,700 nodes may compete for just 500 connections across 4 slots (1.6s). If your transaction can’t squeeze in, it gets stuck.

Now imagine it’s stuck on a node—what happens? If your CU price is too low and the next leader is full, but still sees the transaction, what do you think happens next?

Exactly—data gets sold. High-volume nodes may sell order flow to Searchers for up to $10K/month.

• Meme Narrative and Staking-Driven Volume

Right now, the dominant narrative on Solana is Meme coins. These pools are shallow, so users often need to tolerate higher slippage to get filled. This significantly boosts Searcher profits.

Sampling shows some attacks yield $2 profit per trade on Solana—compared to ~$0.10 on Ethereum. That’s a massive difference.

[Image source: https://www.jito.network/zh/stats/]

Secondly, Solana validators earn around 8% APY from staking, a relatively stable figure over the years.

After integrating MEV strategies, the additional APY from MEV can reach approximately 1.5%.

Combined, this means validators running the Jito-Solana client can boost their staking returns by 15–30%.

In some cases during market surges, MEV profits can even exceed the base staking yield.

2.4 Why Are Solana Validators Prone to Defecting?

The profits are simply too attractive—while the operating costs are also high, pushing validators to constantly expand sources of revenue.

A validator pays roughly 300–350 SOL per year in voting fees (around $42,000 at a market price of $140), plus $4,200 in hardware costs—not including the dynamic costs of maintaining network bandwidth.

Solana’s heavy node requirements demand at least 24 CPU cores, 256GB of RAM, and 2×1.9TB NVMe SSDs.

Custom Latitude machines, commonly used in the ecosystem (14% of validators use them), cost around $350/month.

As a result, only 458 out of Solana’s 1,323 validators are currently profitable. This is one key reason why the SIMD-0228 proposal was voted down. The proposal aimed to reduce block rewards further, which would inevitably force smaller validators to exit, accelerating irreversible centralization.

Now think: as MEV income increases and core protocol rewards shrink, what’s likely to happen? Let’s explore how competitors outside Jito are reacting to this shift.

3. MEV Competitors on Solana

3.1 Paladin: VIP Front-running and Trade Protection

Current market share: ~5%

Launched in late 2023, and as of March 2025, Paladin claims 205 validator nodes have deployed its client, with 53M SOL staked. Validators using Paladin report around 12.5% increase in returns.

At its core, Paladin is a forked and modified version of the Jito-Solana client.

Its key features include:

1. P3 Priority Port: Enables the block-producing leader to open a fast VIP lane, restoring original FIFO processing order.
2. Detection and Filtering of Sandwich Attacks: Though it seems to reduce validator rewards at first, Paladin validators are compensated via a trust-based mechanism. By avoiding sandwich attacks, they attract direct user transactions, building a trust-based ecosystem and improving long-term yields.
3. Paladin Bot: An open-source high-frequency arbitrage bot that runs locally on validator nodes, activated only when the node becomes a leader. It executes simple, risk-free MEV strategies like DEX arbitrage or CEX/on-chain price gaps, and channels profits directly to the validator.

As of December 3, 2024, Paladin officially deprecated the bot.

3.2 bloXroute: Network Layer Optimization + Private Channel Protection

bloXroute Labs is an infrastructure company that provides Blockchain Data Network (BDN) services, previously accelerating transaction broadcasting and reducing latency on chains like Ethereum. bloXroute does not directly engage in MEV allocation but helps front-running transactions reach the Leader faster by providing faster channels. Unlike Jito/Paladin, bloXroute does not modify the Solana validator client or introduce auction mechanisms for transactions but instead offers faster message channels at the network layer for all nodes. Its main approach is:

• Solana BDN Acceleration: According to official documentation, bloXroute’s Solana BDN reduces block shard propagation delay by 30–50 milliseconds.
• MEV-Protect RPC Service: Similar to Ethereum’s private transaction ports, bloXroute plans to allow users to send private transactions to the Leader via its RPC, preventing third parties from seeing them and safeguarding against front-running or sandwich attacks.

3.3 BlockRazor: Network Layer Optimization + Private Channel Protection

BlockRazor, a newly established MEV infrastructure project in 2024, is primarily team-led by individuals from Asia. It positions itself as an “intent-based network service provider,” planning to offer MEV Protect RPC, high-performance network acceleration, and MEV Builder services across mainstream blockchains.

Scutum MEV Protect RPC: This is BlockRazor’s private transaction gateway service, similar to Flashbots Protect. Users can submit transaction Bundles via the Scutum RPC, and BlockRazor ensures these transactions are not released through the public mempool but sent directly to block producers to avoid front-running or sandwich attacks.

4. Summary

4.1 How to View the MEV Competitive Landscape on Solana?

Just the day before yesterday, a new competitor entered the scene: Warlock Labs, which raised $8 million on March 27, 2025, aiming to reshape on-chain order flows. However, their focus is on the Ethereum track. Their plan is to further provide some proof and register on-chain order flow data to ensure accuracy and responsibility in handling users’ transactions. This aligns with my view: a truly good market will continually see new competitors enter, while a market dominated by a few will limit challengers and create barriers. What kind of market do platform layers aspire to become?

Let’s think more deeply: in this MEV infrastructure, what truly matters? Paladin is built on jito-solana, which means jito can upgrade to a version that no longer supports the so-called P3 channel. This is similar to the “3Q battle” of the past, where the winner was the one who was most needed (clearly, social media), and the same logic applied when WeChat banned sharing of NetEase Cloud Music in WeChat Moments. If there were no larger regulatory machines in place, this exclusionary competitive strategy could be used indefinitely in any track. Today, Paladin’s 5% market share is due to early use of built-in bots to increase validator profits. Although its open-source bots are designed to avoid aggressive tactics (like front-running and sandwiching that harm users), they were still pressured into being phased out by market opinion. Other competitors like bloXroute and BlockRazor take the route of acceleration and privacy channels, where privacy ultimately limits the leader to just one point, to avoid malfeasance and ensure mutual retaliation in case of misconduct.

Acceleration capability is a solid technical strength today, and it is the next focal point in the wallet/Dex market competition. Objectively speaking, Solana’s original client code still has some historical debt, which is why someone had to step in and modify the client, making it more efficient with faster synchronization. Combined with Swqos’s mechanism, validators can improve link stability and success rates. Furthermore, jito’s block engine system is a multi-center system, but even with multiple centers (not fully decentralized), single points of failure can still occur. Since it’s a core upstream component, any failure here essentially equals a Solana downtime. Therefore, to achieve multi-node disaster recovery and speedup, it still needs to undergo system challenges like the ones previously tested. This explains why Binance Wallet’s bugs appear more frequently—many of the historical technical debts remain unresolved.

However, this issue of technical strength will ultimately be solved. Anyone can optimize global multi-node systems, and the leader will position their infrastructure in the most efficient way to ensure that their transactions reach the leader the fastest. They may even implement multicast strategies to redirect different user demands. In the future, the competitive outcome will likely hinge on more refined operations. However, an unsolvable issue remains: market competition squeeze. If jito-solana uses its oligopoly advantage to modify the Bundle priority strategy from 80% to 90%, or even 95%, ordinary users will have to raise their Priority Fees endlessly to compete for the remaining 5% CU space. But when total CU usage is insufficient, this will eventually affect the overall validator rewards. With so many transactions piling up in the unprocessed queue, the incentive for validators to act maliciously becomes stronger. Therefore, unless absolutely necessary, Jito would avoid launching such a competitive mode.

So, why is the ETH market competition more open, while Solana’s is more exclusive? The root cause, I believe, is the lack of a Builder bidding role. ETH allows multiple Builders to create multiple final block sequences, with validators only verifying and choosing which one to select. On the other hand, Solana only has multiple block engines (all from the same company), and the transaction queue provided to validators is a single Bundle (5 transactions). This misses the competitive aspect of having multiple Builders. Objectively speaking, from ETH’s development history, it’s clear that this competition significantly boosts validator rewards while reducing Searcher profits. When Searcher profits decrease, attacks become less frequent, leading to a balance.

In the future, where both technology and market balance out, what will the true competitive advantage be? I believe that once the technological gap is closed, accompanied by competition for talent and investment, and once the centralization and decentralization impact on Solana’s overall ecosystem is resolved, the issue will be solved. Solana has already started discussing multiple Builders and even further, has begun discussions about multiple leaders and random block production. Although multiple leaders mean more people will access your orders, since the actual block producer is a random selection from the multiple concurrent queues, this indirectly achieves competition among multiple Builders. The market impact will follow a similar route as before.

So, the true competitive advantage will shift to the data island of order flow. For example, Jupiter, which controls over 80% of the DEX market, has the largest order flow, and now it will need to balance whether to offer the best prices or randomly select a “lucky goose” to capture additional profits, even at the cost of some brand reputation. The reason they are not diving into MEV infrastructure themselves might be due to the developmental stage of the market—no one can claim to be as invulnerable as traditional giants, so focusing on profit will give competitors the opportunity to overtake. MEV is always a game theory problem. Once it reaches a monopoly position, the monopolist’s validator support will push the infrastructure to offer profits. Any dragon-slaying hero seems inevitably bound to become the dragon itself, a combination of the hero and the dragon. Of course, you might argue that Jito’s infrastructure was built specifically for MEV, so how could it be a dragon-slaying hero?

4.2. Jito’s Contribution and Drawbacks to Solana

Much of the previous discussion focused on Jito’s drawbacks, but does Jito have any contributions? Objectively speaking, Jito does have contributions. When I started looking at Solana three years ago, I dismissed it (okay, I admit I was too vocal back then), but the reason for such analysis was its high downtime rate.

Why was there such a high downtime rate? On one hand, there were too many issues in the early code, and later it was found that money could solve most of the problems (machine configurations were continually upgraded). On the other hand, the FIFO strategy played a role: when a high-profit trade appeared on-chain, even if it was just a backrun attack, the closer one was to the transaction, the higher the profit. Obviously, every Searcher would build their own infrastructure to send transactions to the leader as quickly as possible, so early leaders were often flooded with attacks. The emergence of the blockengine created a bidding process. Once you saw the profit, you would immediately bid for it, and traffic would be diverted. This auction also had a feature to intercept failed transactions. If your transaction conflicted with someone else’s, and the other’s price was higher, then because both Searchers were bidding for the same transaction, there would be a storage conflict. If you couldn’t win the bid, the blockengine would directly reject your transaction, forcing you to raise your bid and continue the auction (there is also some randomness to this rejection, which may make you think you must further bid, like a “big data killing familiarity” scenario). Of course, you might ask, why do we still see many failed transactions on Solana? That’s because the blockengine is multi-center. With the 400ms block speed, it struggles to quickly synchronize data, leading to auction errors between different blockengines. Therefore, I believe Jito has made a contribution, as it significantly reduced Solana’s downtime rate.

Beyond downtime, its bundling of transactions has introduced multiple use cases to the market. For a market to thrive, it must serve market makers well. Solana’s most explosive sector has been the Meme market, which relies heavily on the launch group that needs to “subtly” start collecting low-priced tokens as soon as they launch. This is a highly targeted scenario—if the market operator can’t collect enough profitable low-priced tokens, they might abandon the rally and restart the launch. This is a lose-lose situation because the operator will waste an entire launch. Additionally, other use cases, like decentralized exchanges (DEX), now trust Jito-Solana not to blatantly sell data like before. So for high-value transactions, users are encouraged to give a tip to use the fast route through the blockengine, which occupies 80% of the CU processing queue, speeding up the transaction and avoiding being front-run.

Jito has also increased the staking rewards for Solana validators, improving the overall level of decentralization. As previously analyzed, Solana’s staking rewards are around 8% annualized, and with Jito’s MEV tip rewards, this can reach around 10%, which is a good margin. Among Solana’s 1,323 validators, only 458 are profitable. The rest are not entirely unprofitable (otherwise, who would keep doing it?). Some are either acting maliciously or have indirect motives, such as speeding up Swqos. Essentially, the statistics mentioned are based on staking rewards, not including MEV rewards. It is because of Jito’s existence that the remaining 800 validators have become profitable, preventing Solana from being overly centralized. Therefore, from an overall perspective, Jito-Solana does deserve credit. At least at this point, it hasn’t fully adopted an exclusionary competition strategy, which still leaves room for third-party entrants.

4.3. How Will the MEV Landscape Evolve in the Future?

As previously mentioned, there are several key points. I believe that while it currently appears to be a market dominated by one major player with several strong competitors, there is actually an opportunity lurking in the background.

First, MEV profits on Solana are generally higher (around $2, compared to Ethereum’s $0.1), which will continue to drive the meme trend and generate eternal trading opportunities in various narrative scenarios. This means new Searchers will enter the market. Although the higher cost of obtaining order flow on Solana has deterred smaller players, competition among the larger players will intensify as profits increase, leading to greater investments.

Second, there is significant opposition on Solana against MEV infrastructure, which has forced Jito to shut down its data selling channels and Paladin to remove its built-in bot functionality. In proposals like SIMD-228 and the already approved SIMD-96, a portion of the rewards that validators used to receive (which was originally the sum of base fees and Priority Fees) is now being burned. Only half of the base fee will be burned, indirectly increasing the rewards for validators processing normal user transactions, thus enhancing their motivation to counter Jito’s reduction in the weight of ordinary transactions. New proposals continue to participate in the macro decision-making process for Solana.

Third, the overall profit potential of MEV is significant. For example, in October last year, Jito Labs’ fee revenue was $78.92 million, double the record of $39.45 million set in May. This was higher than traditional DeFi protocols like Lido and Uniswap. Even if Jito has to share the profits with validators, the overall profit margin—representing the minimum loss for users—remains large. The greater the loss, the stronger the motivation, and users’ expectations for reliable services can be quantified. This is where BlockRazor and bloXroute come in with opportunities.

Additionally, what excites me are some more cutting-edge explorations:

• Privacy-based transactions: Threshold Encryption, Delayed Encryption, and SGX Encryption are all focused on encrypting transaction information and applying decryption conditions, which could involve time-locks, multi-signatures, or trusted hardware models.
  Fair trading: Concepts like Fair Sequencing Service (FSS), MEV Auctions, MEV-Share, and MEV-Blocker are designed to address the trade-off between no profits and sharing profits, allowing users to decide the cost they are willing to pay to achieve relative fairness in transactions.
• Protocol-level improvements: PBS (Proposer-Builder Separation) is currently a proposal from the Ethereum Foundation, but its implementation through MEV-boost has created a separation. In the future, such core mechanisms will likely become part of Ethereum’s own protocol.

Many of these ideas have already been proposed within Ethereum, but due to compatibility differences, they have not yet emerged into the user’s view. However, these are areas Solana can also learn from and adopt.

5. Final Thoughts

The endgame of competition is often not surpassed by efforts within the same track. It won’t be the next Jito that takes down Jito (as it too has its merits and flaws), but a completely new form of application. In my earlier research, “UniswapX Protocol Analysis“ I summarized the operational flow and profit sources of UniswapX, aiming to fully outline the specific returns from MEV. After all, this is the source it fights against and distributes rewards to users (essentially sacrificing the real-time nature of transactions in exchange for better swap prices). Similarly, order-book based exchanges (even decentralized exchanges) are also powerful tools against MEV. As computational power continues to improve and daily transactions increase, the AMM mechanism and the corresponding MEV attack scenarios will gradually dissipate. However, the challenges that order-books face are not smaller than those of MEV.

From the recent turbulence of Hyperliquid, it is evident that aside from the concerns around centralization, as the web3 ecosystem moves towards regulatory compliance, the players at the table have already donned their suits and entered international halls. At this point, compliance is an all-encompassing sword, as it ultimately stands on the side of the users.

This article, sprawling across thousands of words, is logically and data-driven with insights drawn from various articles. Thanks to the research that has contributed to the development of the industry!

Disclaimer:

1. This article is reprinted from [Fourteen Jun], and the copyright belongs to the original author [Fourteen Jun]. If you have any objections to the reprint, please contact the Gate Learn team, which will handle it as soon as possible according to relevant procedures.

2. Disclaimer: The views and opinions expressed in this article represent only the author’s personal views and do not constitute any investment advice.

3. Other language versions of the article are translated by the Gate Learn team. The translated article may not be copied, distributed or plagiarized without mentioning Gate.io.