The first version of Magic Eden went live before anyone had time to pretend it was finished. Jack Lu and three friends had day jobs, a Solana NFT idea and a market moving at an unreasonable speed. They worked at night, hunted down creators and, by Lu's account, got a site online in about fourteen days. One account of the launch puts its first day at roughly $50,000 in trading volume. It was an answer delivered faster than a business plan could have been typeset.
Five years later, Lu is living with a slower answer. The company that grew out of that improvised launch became a large NFT marketplace, spread to other blockchains and raised money at a $1.6 billion valuation in 2022. The NFT market later contracted. In 2026, Magic Eden wound down its Bitcoin and Ethereum-compatible NFT marketplaces and directed more energy toward Dicey, a crypto gaming business. It is a second beginning with considerably more history attached.
Lu has spoken about how difficult it was to accept that the old market might not return. That admission gives this story its shape. He knew how to start something when demand was arriving in a rush. Now he has to decide what to keep, what to leave behind and how to learn again in public.
The friendship before the company
Long before there was a marketplace, there was a journey to maths tutoring in Melbourne. Lu met Sidney Zhang when they were teenagers heading to the same class. Both had come to Australia from China with their families. Lu was five when his parents moved. He has described the sacrifice of leaving established careers behind and the example of a father willing to try a new country in midlife.
The two boys became friends, then took different routes into technology. Zhang moved to the United States and tried to build a Bitcoin startup called HelloBlock. Lu visited him in Boston, considered joining and decided the market was too early. He went back to Australia. The choice stayed with him: in a later interview he described regret at passing on the uncertain path that his friend had chosen.
He studied commerce and law at Monash University, then worked at Boston Consulting Group. A move to San Francisco led to product management at Google. Lu has described working on Enhanced Conversions there, an experience that put him close to the machinery of consumer products at scale. FTX came next. In Hong Kong, he worked in partnerships and corporate development and watched Solana's ecosystem develop while NFTs were turning digital collections into a crowded public spectacle.
In September 2021 he messaged Zhang about the opportunity he thought he saw: a marketplace for Solana NFTs that was simple for creators and collectors to use. Zhang was ready for another attempt at a startup. They brought in Zhuoxun Yin, whose background included crypto exchanges, and Zhuojie Zhou, an engineer Zhang knew from earlier work. The four had different specialties and a shared impatience with waiting.
A storefront built at night
Early Magic Eden was a practical exercise in meeting a market where it stood. The founders gathered NFT contract addresses, contacted artists and assembled a product while still employed elsewhere. Some initially used pseudonyms. Their work was small enough for each new creator to matter. Once the site launched in September 2021, the response came quickly: the first month brought about 200 listed collections, and the founders left their jobs to work on it full time.
Lu later explained the timing without false modesty. Ethereum NFTs had made the idea familiar, and Solana was gaining attention just as the new marketplace arrived. But the team also made a choice that shaped its growth. It did more than offer a place to resell a token. Its Launchpad helped creators bring collections to market. Those creators arrived with their own audiences; each successful launch made the site more useful to the next one.
One early story Lu has told makes the relationship feel less like a sales funnel. The team approached the artist behind an early collection called Cyber Trolls to discuss listing it. The artist turned the conversation around and pitched himself for a job. Magic Eden hired him as an early business development employee. The people it was trying to serve were close enough to the work to help build the company.
Historical launch and fundraising figures. Trading volume is the value of transactions, not company revenue.
The June 2022 financing brought $130 million into Magic Eden. The valuation drew headlines, but the business still had to operate in a market famous for abruptly changing its mind. A marketplace depends on activity, and activity depends on people wanting the things being traded. Lu's team had built a fast, welcoming door. The weather outside the door remained beyond its control.
When the map kept changing
Magic Eden moved beyond its Solana beginnings. Ethereum and Polygon came into view, followed by Bitcoin assets including Ordinals. The company built wallet and mobile products. The logic was visible: collectors do not necessarily stay loyal to a single chain, and a marketplace that follows them can remain part of their routine. Lu argued for an experience that felt as quick and familiar as an ordinary consumer app, even when the underlying technology was complicated.
Expansion created its own burden. Each supported chain and product needed engineering, maintenance and attention. Meanwhile the broader NFT cycle moved through sharp rises and long quieter stretches. Magic Eden also dealt with layoffs in 2023. The 2022 valuation was a moment in time, not a standing measurement of what the company or its market would become.
“The user experience can’t suffer.”Jack Lu, on building a usable crypto product
By 2024, Lu was still discussing wider ambitions for the platform, including a wallet that crossed chains. He had reasons to pursue them. Bitcoin collectibles had brought new trading, and Magic Eden had already shown it could move when a new category appeared. Yet a larger product menu could not guarantee a larger enduring audience. Lu has since described long periods spent fighting for the last increments of market share. The big gains, he said, came when the team did something different.
The reversal became public in early 2026. Lu announced plans to end Magic Eden's EVM and Bitcoin NFT marketplaces, along with related infrastructure, while retaining its Solana roots and selected collectibles products. He said much of the company's cost was attached to products producing a smaller share of revenue. The strategic choice was stark for people who had come to Magic Eden precisely because it supported those chains.
In an interview published in September 2026, Lu gave the pivot an emotional dimension. He had spent months hoping the NFT market would return. Accepting its smaller scale meant admitting that years of work might not regain their old economic weight. He described giving colleagues a choice about joining another startup-like effort. The group working on Dicey was much smaller than Magic Eden's peak headcount.
The second set of first mistakes
Dicey grew from a thought Lu had formed about how people used crypto products. NFTs offered culture, ownership and community, but in quieter times a good deal of activity looked like speculation. The new business addresses entertainment and wagering directly. That is a different kind of consumer relationship: every product decision touches money, trust and a player's sense of fairness. Lu and his colleagues entered it without experience operating a gambling platform.
They began with a small beta rather than a vast launch. The first group of roughly 200 players included people from the Magic Eden community. One spotted that Dicey's blackjack game felt wrong. The team investigated and found a logic error that kept queens from being dealt. Lu says affected players were refunded. It is the sort of bug that turns a product meeting into a lesson about credibility.
Lu has also tried doing customer work himself, including hosting VIP players, to understand what his team is asking others to deliver. He described a case in which the numbers and rewards appeared right but the player still felt poorly served. The lesson was awkwardly human for a founder who had spent years around market mechanics: service is not entirely reducible to an equation.
He wants Dicey to develop original games, clearer rewards and, eventually, ways to make parts of its activity more transparent on chain. These remain ambitions rather than proof of a finished model. In the September interview, Lu reported about $20 million in monthly deposits after beta; that is his figure, and deposits are not revenue. The live question is whether the team can turn early curiosity into a business that people trust and keep using.
What remains of the first beginning
The second launch has an echo of the first. Once again, there is a rough early product, a handful of users whose reactions matter and founders learning as they build. This time there is also a public history. Magic Eden's first customers watched the company expand, and some watched it retreat from products they used. Lu cannot arrive at Dicey as an unknown young founder with only possibilities ahead of him.
That may be why his most revealing story is still the old friendship. A teenage maths class led to a missed Bitcoin opportunity, years in corporate technology and, eventually, a company built with the same friend. Lu's parents had started over in another country. He followed that example by moving to the United States, then by leaving established jobs for a tiny marketplace. Now he is testing how much of the instinct survives when the thing to leave behind is his own success.
The two-week site showed that he could recognize a moment and move. The slower decision of 2026 asks for something else: the patience to hear a market's changed answer and the nerve to build with it. Nobody gets to learn that lesson from a valuation headline. Lu is learning it from the people on the other side of the screen.