Marathon Fusion’s Tiny Gold Output Leaves Bitcoin’s 21M Rule Unchanged

Marathon Fusion says its neutron-driven process produced stable gold at a rate 10 times CERN’s record, but the company has not disclosed the output and still describes the quantity as microscopic. The U.S. startup’s research paper says high-energy neutrons from a deuterium-tritium fusion reaction irradiated mercury-198, creating a radioactive isotope that decayed into stable gold-197.

The paper describes a scientifically plausible nuclear-transmutation process, not a commercially meaningful increase in supply. Marathon still must build a fusion reactor, scale production of the enriched mercury used in the reaction and show that gold recovery can cover reactor and processing costs. The resulting “infinite mint bug” question remains hypothetical because no industrial output has been shown.

Does fusion-made gold strengthen Bitcoin’s case?

Bitcoin’s protocol caps supply at 21 million coins, leaving its programmed scarcity unaffected by a new source of gold. Coinbase chief executive Brian Armstrong compared net-new gold produced through “alchemy” with the lab-grown diamond industry, then said: “But you can’t invent anything that will increase Bitcoin supply. There will only ever be 21M.”

University of Basel professor Fabian Schär said a substantial increase in available supply would place downward pressure on an asset’s market price, other things equal. He extended that logic to gold only if fusion production became inexpensive at scale. For now, extra gold cannot loosen Bitcoin’s own supply cap.

Can Marathon produce gold at commercial scale?

Marathon’s 10-times comparison came against the ALICE experiment at CERN, the European particle physics laboratory, which produced an estimated 29 picograms of gold nuclei during its 2015-2018 run. Collisions destroyed those nuclei after fractions of a second.

Research engineer Jake Harter said Marathon’s present output remains microscopic and that the company hopes to reach macroscopic milligram quantities in 2027. The research paper’s larger figures project a future plant rather than measure the completed experiment.

• Modeled annual output: About 2 tonnes of gold per gigawatt of thermal fusion power.
• Mine-supply benchmark: Global mines produced 3,672 tonnes in 2025, according to the World Gold Council, making one future fusion plant’s modeled output modest by comparison.

Reaching those targets requires a functioning fusion reactor and greater production of the enriched mercury used to produce gold. Harter said the company also plans studies of materials compatibility, mercury pumping and gold extraction. The nuclear process is plausible, but its projected economics remain unproven.

Could fusion gold reshape the gold market?

Marathon co-founder and chief executive Kyle Schiller said jewelers have shown a striking level of interest, with most expected gold likely to serve as a store of value. Metals Focus managing director Nikos Kavalis said the sector is nowhere near production significance and that miners should keep their jobs for the foreseeable future.

Wholesale prices for lab-grown diamonds have fallen more than 90% since 2018 as production scaled. Their chemistry matches mined diamonds at a fraction of the price, unlike substitutes such as cubic zirconia. eToro market analyst Nagham Hassan said fusion gold’s economics differ because it requires complex infrastructure that does not exist commercially.

Kavalis said: “There is no demonstration that the process is commercially viable. There are theoretical revenue projections, but nothing about what it will cost to run the reactors.” That leaves the business case dependent on costs the company has not disclosed.

The gold would also remain radioactive. Harter said it would need storage for a bit over a decade before reaching banana-level radioactivity, though he does not consider that a major obstacle because investors generally store gold. Fusion gold is not yet a material supply competitor to mining.

First fusion reactors are estimated at least a decade away, leaving the work as a potentially useful experiment rather than an imminent threat to gold markets or Bitcoin demand. The next checkpoint is Marathon’s 2027 target of macroscopic milligram output; reactor economics, mercury scaling and extraction must follow before fusion gold could alter the market structure underpinning Bitcoin’s digital-gold comparison.