Direct answer: yes — gold can genuinely be made in a laboratory. CERN proved it in 2025, turning lead into gold in microscopic amounts. On October 21, 2026, San Francisco startup Marathon Fusion announced it had gone further: turning mercury into stable gold using fusion neutrons, at 10 times CERN's rate. But that second claim is company news, not verified science. No yield, no purity, and no production cost has been disclosed, and no independent laboratory has confirmed the result. So the honest verdict is this: lab-made gold is real physics, but a gold factory is not here, and the gold market is not about to collapse.
One date correction first, because viral posts are getting it wrong: Marathon's announcement came on October 21, 2026 — nine days before this article, not within the last five days. The story is real and recent, but it is not quite as fresh as the viral version claims.
What Marathon Fusion actually announced
Marathon Fusion is a San Francisco fusion-energy startup founded in 2023. On October 21, 2026, it posted on X that it had demonstrated "high-rate alchemy": turning mercury into real, stable gold, beating CERN's record by a factor of 10. According to later reporting, the demonstration itself took place in September 2026, with the announcement following in October.
The claim matters because of one word: stable. The gold CERN made at the Large Hadron Collider was gold-203, an unstable isotope that survives only a tiny fraction of a second before decaying away. Marathon says its process produces gold-197 — the only stable isotope of gold, the same kind found in jewellery and central-bank vaults. If that is right, the gold it makes would stay gold. That is the difference between a physics curiosity and something that could, in principle, one day be collected and sold.
Marathon's own scientists are the first to say that gold is not their main prize. The company is building a fusion reactor, and its business idea is to make money from the neutrons its reactor produces — selling medical isotopes and, potentially, gold to help pay for expensive fusion plants. It says it has already made copper-64 and copper-67, isotopes used to image and treat cancer, by aiming the same neutrons at zinc. Gold, in this picture, is a by-product that proves the neutron source works, not a plan to flood the bullion market.
How mercury becomes gold
The physics is straightforward, and it has been understood for more than a century: an element is defined by the number of protons in its nucleus. Mercury has 80 protons; gold has 79. Remove one proton from a mercury nucleus and you get gold. The hard part is doing it without smashing the nucleus apart entirely.
Marathon's route, as described in its 2025 paper, starts with mercury-198, a stable isotope of mercury. Fast neutrons from a deuterium–tritium fusion reaction — each carrying about 14 million electron volts of energy — strike the mercury-198. The impact knocks two neutrons out of the nucleus, leaving mercury-197. That isotope is radioactive, with a half-life of roughly 64 hours, and it decays into gold-197, the stable form. Wait a few days, and the mercury-197 in the sample has quietly become gold.
There is a catch hidden in that chain. The starting material is not ordinary mercury off the shelf: the process needs mercury enriched in the mercury-198 isotope, and producing that enriched material is itself difficult and expensive. The gold also comes out of an irradiated sample, which raises questions about residual radioactivity — more on that below.
CERN vs Marathon: what each one actually did
Putting the two results side by side shows why Marathon's announcement got attention — and why it still needs proof.
| CERN ALICE (May 2025) | Marathon Fusion (October 2026) | |
|---|---|---|
| Starting material | Lead-208 nuclei | Mercury-198 |
| Method | Near-miss lead collisions at the Large Hadron Collider; photons knock protons out of the nucleus | Fusion neutrons strike mercury-198, producing mercury-197, which decays to gold-197 |
| Gold produced | Gold-203, unstable, survives a tiny fraction of a second | Gold-197 claimed, stable — the same isotope as mined gold |
| Amount | About 29 picograms over four years of LHC running (2015–2018); roughly 86 billion nuclei | Not disclosed |
| Peak rate | About 89,000 gold nuclei per second | Claimed to be 10 times CERN's rate — basis of comparison not explained |
| Evidence published | Peer-reviewed paper in Physical Review C, plus a CERN announcement | Company post on X, gamma-ray spectra, and a video; no peer-reviewed paper on the demonstration |
| Independent verification | Yes — a CERN collaboration result, published and scrutinised | None reported so far |
The contrast in the last two rows is the whole story. CERN's result is established science: a published, peer-reviewed measurement by a large collaboration. Marathon's is a company announcement. That does not make it false — startups often announce results before publishing them — but it does mean the claim currently sits in a different category of evidence, and it should be reported that way.
What is missing: the unverified parts
As of late October 2026, publicly available reporting has not established several basic facts about Marathon's demonstration:
- How much gold was made. The company has not said whether it produced a visible amount, a measurable trace, or something in between. Without a yield, "10 times CERN's rate" cannot be checked — and CERN's rate was already vanishingly small, so ten times a microscopic number is still microscopic.
- How pure it is, and how it was recovered. Making gold atoms inside an irradiated mercury sample is one step; separating out pure, verifiable gold is another. No recovery method or purity figure has been disclosed.
- What it cost. No production cost per gram has been published, so nobody can say whether the process is remotely economic — and the economics are the entire question for a "gold factory" story.
- Why the comparison to CERN is valid. CERN's benchmark is a peak rate of nuclei per second in a collider. Marathon has not explained exactly what it measured to claim a factor of 10, so the comparison cannot yet be audited.
- Independent confirmation. No outside laboratory has replicated the result, and the company's 2025 paper describing the process is a preprint, not a peer-reviewed publication. The evidence shown so far consists of gamma-ray spectra the company says support its isotope identification.
Sceptics have been blunt. One commentator compared the announcement — a paper that is not peer-reviewed plus a short video — to a junior mining company promoting a prospect without assay results, and noted that the company has not said whether it produced any gold at all. That is harsh, but it captures the standard the claim will be judged by: extraordinary claims need independent evidence, and a social-media post is not that evidence.
What it would take to actually matter
Even if the demonstration is fully confirmed, the road from a laboratory result to anything that affects the gold market is long. The numbers show why.
Marathon's 2025 modelling estimates that a proposed future plant could produce around two tonnes of gold per year for every gigawatt of thermal fusion power — roughly five tonnes a year from a plant generating one gigawatt of electricity. Set that against global mine production of about 3,672 tonnes in 2025: one gigawatt-scale plant would add roughly one-seventh of one percent to world supply. It would take hundreds of such plants to move the market, and no commercial fusion power plant exists yet at all.
The scale gap is the immediate problem. Marathon's current reactor reportedly runs at around 1 kilowatt. A gigawatt is a million times larger. Bridging that gap is not a matter of turning a dial; it is decades of fusion engineering, and fusion power itself remains unproven at commercial scale.
Then there are the practical hurdles. Enriched mercury-198 is difficult and expensive to produce. Gold made by neutron irradiation can carry residual radioactivity, and Marathon's own simulations suggest the material could need up to 18 years of storage before it meets stricter safety standards for commercial sale. And the process must compete with mining on cost — a bar that no laboratory transmutation has ever come close to clearing. Marathon's route is designed to be cheaper, but until it publishes a cost, "cheaper than CERN" tells us nothing about "cheaper than a mine".
That is why market analysts are calm. One analyst assessment in late October 2026 concluded that lab-made gold is unlikely to disrupt global supply or prices any time soon, because the technology is unproven at scale and the commercial hurdles are large. For investors and gold owners today, the announcement is a non-issue. For physics, it is genuinely interesting. Both things can be true at once.
What happens next
Three things would turn this from a viral claim into established science. First, a peer-reviewed paper with the full data: how much mercury was irradiated, how much gold was recovered, how purity was measured, and how the rate comparison to CERN was calculated. Second, independent replication — another laboratory repeating the experiment and finding gold-197. Third, a credible cost and scale roadmap: what a gram costs at pilot scale, and what a plant would have to look like for the two-tonnes-per-gigawatt model to mean anything.
Watch for those three milestones, in that order. A second company post, a longer video, or a bigger claimed multiple of CERN's rate would not change the picture. A published paper with numbers would. Until then, the fair summary is one sentence: scientists can make gold in a lab, a startup says it has found a better way to do it, and the world is waiting for the proof.
Frequently asked questions
Can scientists really make gold in a lab?
Yes. CERN's ALICE experiment measured the conversion of lead into gold at the Large Hadron Collider in 2025, and nuclear transmutation — changing one element into another by altering its nucleus — has been understood for over a century. The gold CERN made was microscopic in amount and unstable, but the physics is confirmed.
Did Marathon Fusion really turn mercury into gold?
That is the company's claim, announced on October 21, 2026. It says it converted mercury-198 into stable gold-197 using fusion neutrons at 10 times CERN's rate. The claim has not been independently verified, and the company has not disclosed how much gold it made, its purity, or its cost. Treat it as an unverified company announcement, not established science.
Will lab-made gold crash the gold price?
Not on anything visible today. Even Marathon's own model for a future gigawatt-scale plant suggests around five tonnes a year, against about 3,672 tonnes mined globally in 2025. No commercial fusion plant exists, the current test reactor is around 1 kilowatt, and the gold may need years of storage before sale. Market analysts see no near-term threat to supply or prices.
Why is Marathon's gold different from CERN's gold?
CERN's gold was gold-203, an unstable isotope that decays within a tiny fraction of a second — it cannot be collected or sold. Marathon claims gold-197, the only stable gold isotope, identical to mined gold. That is why the claim attracted attention: stable gold could in principle be kept. Whether the claim holds up is still unproven.
Is this the same as the old alchemists trying to make gold?
The goal is the same, but the method is real physics rather than chemistry. Alchemists tried chemical tricks, which cannot change one element into another. Nuclear transmutation changes the nucleus itself — removing a proton from mercury genuinely leaves gold. The alchemists' dream turned out to be possible; it just turned out to be spectacularly uneconomic.
Sources
- CERN: ALICE detects the conversion of lead into gold at the Large Hadron Collider (May 8, 2025) — https://home.cern/alice-detects-conversion-lead-gold-lhc/
- ALICE Collaboration: Proton emission in ultraperipheral Pb-Pb collisions at 5.02 TeV, Physical Review C 111, 054906 (2025) — https://arxiv.org/abs/2411.07058
- Marathon Fusion announcement on X (October 21, 2026), as reported by Euronews and others — https://www.tradingview.com/news/benzinga:18f391421094b:0-another-golden-asteroid-story-why-lab-made-gold-won-t-crash-bullion/
- Marathon Fusion mercury-to-gold claim: reporting and analysis — https://www.tradejini.com/blogs/can-labgrown-gold-change-the-gold-market
- Marathon Fusion's proposed route and 2025 modelling — https://www.bitveen.com/make-gold-in-a-lab-mercury-to-gold/
- Market assessment: lab-made gold unlikely to disrupt supply or prices — https://mid-east.info/lab-made-gold-unlikely-to-disrupt-global-supply-or-prices-anytime-soon-2/
Be the first to share a useful question or thought.