Jack Levin Interview, Part 2: The XNT Burn, Alpenglow, Capybara and XCORE's Bonding Curve

Twelve days after his first long conversation with X1 Report, X1 founder Jack Levin came back to talk about what is happening now: the 85 million XNT burn, Solana's Alpenglow upgrade, the Capybara delegation floor and XCORE. He spent most of the hour sharing his screen, showing two things the community had only heard about in pieces. He spoke with Owl of Atena on October 22, 2026. The full, uncut video is above.

"One of my favorite topics"

Ask Levin about a consensus upgrade and he lights up. Before he got to X1's burn, he wanted to explain what Solana had just done, because, as he put it, optimization and performance are "one of my favorite topics."

Under the old design, he explained, every Solana validator sends its own vote to the block leader. The leader has to check every signature, look up each voter's stake and work out whether two thirds of the stake agree. That is the ceiling.

“The problem with this architecture is that your network cannot grow to, let's say, above 10,000 validators. All of those validators yell at the leader to look at their data. Even if a validator has 2 XNT staked and its vote is really useless, the leader listens to it anyway, and it overloads the leader with information and traffic.”

Alpenglow, he said, works more like Ethereum. Validators aggregate their votes among themselves with BLS signatures, and the leader receives one signature instead of thousands. The day it switched on, the vote traffic simply vanished from block explorers.

“If you looked at Solscan, there were three or four thousand vote transactions, and boom, they all disappeared, because now they're hidden behind this extra protocol. The leader only looks at one signature, not 10,000 signatures.”

Levin has tested BLS aggregation himself. Building the aggregate takes a little time, he said, but verifying it takes "a couple of milliseconds." Finality drops from about 12 seconds to about 150 milliseconds. He added a caveat that only an engineer would: finality is not the same as everyone seeing the result. "If you're on the Galapagos Islands or in Antarctica, the information comes to you much slower."

And a personal note: "Hopefully Solana doesn't crash, because I have a lot of Solana tokens."

Editor's note: the voting part of Alpenglow is called Votor, and it went live on Solana mainnet on October 21. Votes are no longer on-chain transactions. Solana's official figures put the old finality at about 12.8 seconds and the new target at about 150 milliseconds. More in our burn article.

Why 85 million XNT

Solana's finality got roughly 85 times faster, so X1 will burn 85 million XNT. But Levin said the idea is older than the tribute. He borrowed it from Binance, which burns BNB according to its trading activity, and had wanted to do something similar since before X1 launched.

“Our fees are very, very low and remain very low, because the activity on X1 is obviously not as hot as Binance or Solana. So we can't really tap into the activity metric. And so I thought, why not just go and burn all of the inflation that happened this last year, from October 6 to basically now.”

His explanation of why it matters to holders is arithmetic. If you own 1% of the supply and the supply shrinks, you own more than 1%, without buying anything. He gave an extreme example to make it obvious: burn half the supply and every holder's 1% becomes 2%.

That only works, he said, because X1 holds so much of its own supply. Ethereum's foundation holds about 1% or less, so a burn would barely register. Solana's foundation doesn't hold much either.

“We can do it because we have a lot of it, and we would rather have all of the investors have more of a supply percentage.”

The coins will go from a Foundation multisig to Solana's incinerator address, which X1 inherits. "It spells incinerator, with some letters converted to numbers so that it actually works," Levin said. "That address has no private key. So nobody can get those funds out."

Editor's note: we checked X1 mainnet when this was published. Total supply read 1,081,097,642 XNT and the incinerator, 1nc1nerator11111111111111111111111111111111, held 0 XNT, so the burn had not been executed yet. Anything sent to the incinerator is destroyed at the end of the block, so the proof will be the transaction itself and a drop of 85 million in total supply.

Where the 85 million comes from

The burn will come out of the Foundation Treasury wallet, not out of anything held by the public. Almost all XNT is non-circulating, held by X1 itself. This is how that non-circulating supply breaks down before the burn:

Taking 85 million out of the Foundation Treasury's 501.62M removes about 17% of that wallet and leaves it near 416.6M XNT. Total supply falls from about 1.081 billion to about 996 million. Because the coins come from the treasury, the circulating supply, the part the public holds, does not change. What changes is everyone's share of the total. We track every holder group, the circulating float and the staking split on our XNT price page.

"It ends up being the same thing as if it was given"

The question waiting for him had come up in Telegram that morning: why burn the coins instead of handing them to the community? Levin's answer was that the two are the same thing, and the burn tells a better story.

“Your XNT, even though the number did not change, is more valuable because of the supply reduction. So it's the same thing, but from the marketing perspective it looks better that your supply is not growing and there's no inflation.”

“The single transaction affects the value of XNT in everybody's wallets, including my wallet also. My allocation was x%, but now it's x% plus another 8% of my allocation.”

It benefits everyone equally, he said, "even those people who haven't purchased XNT yet," because every XNT bought after the burn is a slightly larger slice of the total. He thought XNT had already moved up by maybe 10% on the announcement.

A burn every quarter

This will not be a one-off. "I think we probably will be burning once a quarter," Levin said. The goal is to hold total supply around one billion. The 85 million takes it below that line, but only "for like a month maybe," before inflation, now "less than 7%" and falling every year, brings it back.

“Ideally, if we stay at about 1 billion supply and burn every quarter to hit the 1 billion supply target, then we don't have the inflation eroding the value of the token.”

He pointed at Solana as the cautionary tale. Inflation, he said, was "a silly argument, but it is kind of somewhat true": SOL was around $250 three years ago and touched $250 again recently, yet those two prices are not the same market cap. With constant inflation, a token has to fall 5 to 10% a year just to stand still.

Could the burn become automatic, like Binance's or Hyperliquid's buybacks from exchange revenue? Not yet. "If there's some juicy activity happening, maybe we can connect to that. But for now, we'll just do it manually, because it's very simple." X1 wants fees close to zero, he added, just high enough that spam isn't free.

Capybara: "Some people are happy and some people are crying"

Capybara raises the minimum stake a validator needs to receive Foundation delegation. As we reported on October 20, the floor moves to 20,000 XNT from Tuesday, October 27. Levin did not pretend it was popular with everyone.

His case is about who the Foundation's generosity was really feeding. With delegation, he said, yields run "not 7 or 8%, but more like 300%." That attracted operators whose only plan was to collect it.

“There's a lot of very, very smart people whose sole reason to run their validators is to farm the foundation delegation. So that being the case, it's really net negative for the network, because they're farming and selling, farming and selling, farming and selling.”

The people he wants to reward are the "hardcore old-timers, OGs, people who love X1," who are "putting their money where their mouth is." And the farmers still have a way in:

“If the farmers want to make more XNT, well, they can buy some, they can buy and stake, and they can earn extra 300% yield that no blockchain is giving to anybody. We're very generous with our yield, because we want you guys to have the yield. But we also want you guys to be the buyers.”

The slippage problem

Before XCORE, Levin wanted to show the problem that comes first. He opened the newly redesigned XDEX and pointed at the XNT pool: about $8,000 of liquidity. Then he typed in buy orders. $5,000 pushed the price to about 45 cents. $6,500 moved it to 50 cents. At $10,000, the buyer was paying 75 cents per XNT, in a market trading near 20 cents.

The pool is thin for a reason, he said. Anyone adding liquidity at 20 cents is volunteering to sell at 20 cents, and nobody wants to. "People want to sell at 85 cents and a dollar, or higher, $10."

“Every day I get people DMing me and messaging me and asking me, can we get some XNT over the counter? And I just tell them, just go to XDEX, or go to somebody who has XNT that they want to sell, and negotiate the rate.”

Editor's note: these are the numbers from Levin's screen during the call. Pool depth and price move constantly; live figures are on our XNT price page.

An exchange counter for XNT

His answer is a program that works like the currency kiosk at an airport. It sells XNT at one fixed rate and buys it back at a lower one. There is no curve and no price impact, only the spread.

“This is basically a liquidity providing facility, and it works similar to how a currency exchange works. The exchange right now is set to $1 to buy XNT and 90 cents to sell XNT, but it has no slippage, it just has the rate differential. Those numbers can be set by us to be whatever we like.”

In the demo he bought 10,000 XNT for $10,000, then sold it straight back for 9,000 USDC. The program keeps the 10% haircut, and Levin has plans for it: every hour, the retained USDC goes to XDEX, buys XNT, and "potentially" burns it. That buying pulls the XDEX price toward the counter's rate.

The real rate will not be a dollar, he said. It will sit a little above XDEX: if XDEX is at 20 cents, perhaps 30 or 35 cents to buy, about what a large order on XDEX would cost anyway, with sellers paid around 30. The few cents in between fund the buy orders on XDEX. Arbitrage, he argued, is welcome. XDEX's pool is so small that a single trader could close the gap within an hour, and once it is closed, liquidity providers can add to XDEX without selling cheap.

Owl suggested selling in price steps, so nobody could buy an unlimited amount at the lowest rate. Levin rejected it.

“If we were to do it that way, that will be an equivalent of slippage, and we really don't want slippage. I don't think we should limit it at all. If people want to buy $10 million worth of XNT, well, that's great.”

Setting the rate, he joked, makes someone on the team the chain's central banker.

“The team and I were joking that somebody has to be Jerome Powell, because he sets the rates based on the economic activity. Well, based on our economic activity, let's set the rates. People have a desire to buy it. They just don't want to buy it with large slippage.”

Is this the long-promised terminal? "We can call it terminal," he said. What he had in mind before was more complicated; this is the simplified version, and he wants one on Solana too, so wXNT buyers there avoid the same thin pools. Of all the pieces he showed, it is the closest to ready: "The terminal program is easier to launch because it's already done. Maybe early November will go live."

Editor's note: the rates, the 10% haircut and the 30 to 35 cent example are illustrations from the call, not launch settings, and early November is Levin's hope, not a confirmed date. You can try the simulator at wxnt-sale-sim.vercel.app/xnt-swap. It runs on fake balances in your browser, and a second tab sends trades to an xnt_swap program on X1 testnet with test tokens that have no value.

XCORE, version two

Then XCORE. Levin's pitch for it has not changed: a Solana token with a fixed supply of 1 billion, built for traders, that pays its holders in wXNT. What has changed is how it would reach them.

First come free claims. About 200 million XCORE would go to people who held XEN, HEX, VMPX and other assets in the XEN family at a snapshot taken October 1. The Ethereum side of the snapshot, he said, is Ethereum only. Whether stakers count as well as liquid holders is still open: "I haven't decided on stakers yet. Maybe just liquid holders at that block, but we'll see." There would be a small flat claim fee, both to stop abuse and to raise a little USDC.

Editor's note: the public claim demo uses a sample holder list and a mock claim; no real XCORE exists. On October 22 it split the 200,000,000 pool as XEN 40%, HEX 15%, XNM 25%, solXEN 12% and XENCAT 8%, with rewards curved so large holders get less per token than small ones. A separate claims table lists 500 sample addresses built from a test snapshot, not the October 1 one, with HEX capped at 100,000 XCORE per claim and pool, contract and burn wallets excluded. Its VMPX section shows VMPX holders sharing a fixed 25,000,000 XCORE pool, pro rata to their October 1 holdings. Levin said the allocation numbers still have to be decided, so all of this is sample data.

After a claim window of "maybe a week, maybe two weeks, maybe 10 days, I don't know," the rest would go on sale on a bonding curve that only goes one way.

“Every time I make a buy, you see how this number is getting smaller. That's because the curve is rising. So the more people buy it, the less supply there's left and the rate is moving higher. The only difference between this and an AMM is that you cannot sell back to the pool. You can only buy from the pool.”

In his demo the slippage stayed around 0.6%, while each $10,000 bought less XCORE than the last. When the curve sells out, the token "graduates", in his words "similar to pump.fun": part of the USDC raised and part of the XCORE held back seed a Raydium pool, and the keys are burned. Only then can claimers sell, or keep holding for the wXNT.

His rough split: 200 million for claims, "another 200 million probably will be held in the treasury," and "maybe 500, maybe 600 million sold on the curve."

That replaces the plan we described on October 11, a fixed sale of 20,000,000 wXNT at $1 with 45 XCORE bundled per dollar. Asked whether that page was going away, Levin didn't call it dead: "That's the rough design, and that's kind of like where my mind is." The simulator still hosts both designs.

A speculative token for the trenches

Why build a token at all, and why on Solana? Levin laid out three goals: slippage-free liquidity on X1, buy pressure for XDEX through the exchange counter, and a way to get Solana traders to notice X1.

“Solana has maximum liquidity in the world right now, and we want to tap into the liquidity by offering something that can generate yield in wXNT, which is X1 blockchain. That's the best way to introduce X1 to Solana guys as an infrastructure project, plus show them everything else we have. And no better way to do it than with a speculative token that the trenches would really appreciate.”

XCORE would stay on Solana. It can be bridged, he said, but he wants the trading there. A trading fee, which he put at "1, 2, 3%, whatever," would buy wXNT from the XCORE/wXNT pool for holders, and staking XCORE might earn a boost from X1's Foundation yield. "People can speculate on the token. It can go up and down in value, but they can always get some wXNT as the yield."

As for when: XCORE is the harder of the two to launch, with "a lot of moving parts" on Solana and allocations still undecided. He has been sharing the simulator links openly and wants feedback. "Just DM me, and Theo and I will look at it and maybe implement some ideas."

"This is not an interview"

The burn, he said at the end, would probably happen that week. "We'll do a burn there, and we'll announce it too. I'll post the transaction." When Owl thanked him for the interview, he corrected her.

“This is not an interview. This is a friendly stream. I think we have an amazing building culture, and we have AI to help us, and in a bull market we'll be unstoppable.”

What is still open

Levin was clear that much of what he showed is unfinished. As of this conversation:

• XCORE's allocation numbers, the claim split between communities and whether stakers qualify are not decided.
• The sale model is not final. Levin now favours a buy-only bonding curve over the fixed $1 sale, but the curve size, the treasury share, the claim window, the fees and the staking boost are all examples.
• The official XCORE website will carry the final, precise terms. Until then, the simulator pages are demos with fake money.
• The exchange counter's rates and spread will be set at launch. Levin's target is early November.
• The 85,000,000 XNT burn is announced but not yet on-chain.

Read part one for Levin's story from Google to X1, and follow him on X at @mrJackLevin for updates as these plans firm up.