The undisputed leader in advanced geothermal is Fervo, which went public in May via a $6.5-billion IPO. According to data from Currence, Fervo is ahead of its peers on every metric: tech maturity, acreage and offtake, financing, and drilling capabilities. It’s the drilling success in particular that has allowed it to demonstrate the viability of its approach — and raise significant amounts of capital and sign firm offtake agreements as a result.
There’s a large number of startups trailing in Fervo’s wake: Currence is tracking at least 14 other non-incumbents, all of whom are no doubt watching closely to see how Fervo performs on the public market.
The good news for those companies is that there’s a seemingly bottomless demand for clean firm power for AI, and geothermal — particularly advanced geothermal — is moving quickly. Advanced geothermal is an umbrella term that includes multiple approaches to harnessing subterranean heat that go beyond plumbing naturally productive hydrothermal reservoirs, which has been an established though niche industry for decades.
By one estimate, behind-the-meter advanced geothermal could serve more than half of projected U.S. data center load growth over the next decade. The technologies are already attracting interest from hyperscalers like Google and Meta. That said, the sector is still trailing other clean firm options like advanced nuclear, explained BNEF analyst Stephanie Diaz.
“If we look at next generation nuclear technologies…at least as of May, data center operators have announced 43 gigawatts of nuclear projects,” Diaz told Latitude Media. That number includes a lot of announcements and nonbinding agreements, and none of it is guaranteed to come to fruition, she conceded. But it nonetheless far outstrips the next generation geothermal pipeline of 0.6 GW — even though the latter is getting to first-of-a-kind much sooner.
The biggest issue for geothermal is the “geographic mismatch” between where the best, most financeable near-term resources are, and where much of the new data center load is clustering. The lowest-risk geothermal prospects are in the West, while emerging data center hotspots are showing up in the Southeast, Midwest, and Texas. Today, that’s still the case for most advanced geothermal approaches, which promise more geographic flexibility in the long-term.
A pipeline of ‘bragawatts’
That’s not to say that geothermal’s geography and emerging data center load can’t align: XGS, a Texas-based next-gen geothermal startup pursuing a closed-loop system, recently signed a 150-MW agreement to deliver power to the New Mexico grid in support of Meta’s data center operations in the state. It’s a development agreement, rather than a conventional PPA, but initial exploration and engineering work has already begun on the project.
Geothermal’s pipeline of deals may ultimately be much broader than is visible at this point, Diaz said. Lucy Darago, chief commercial officer at XGS agreed: Because geothermal projects are operating on a nearer timeline than advanced nuclear, she said, the signals the industry puts out “have to have real tangibility to them.” There’s definitely “some sensitivity” within the geothermal industry about the massive pipelines that other clean firm technologies are touting, she added, especially given how long their timelines are.
XGS, for example, doesn’t announce agreements that are decades out or aspirational, Darago said. “We don’t want to go public with a deal unless we feel we’ve laid the groundwork with communities, with customers, with utilities, with every stakeholder that could matter in our development to to support the success of that project,” she explained.
While geothermal’s data center “braggawatts” may be lagging behind advanced nuclear’s, the last few years have changed things.
XGS’ unannounced pipeline, for example, has shifted dramatically. Geothermal for decades was just one resource in a broader utility portfolio, Darago explained. Now, a growing share of the XGS pipeline is tied to new large loads: the types of projects where geothermal is the foundational, long-term supply. Early conversations with data center customers often focused on behind-the-meter or islanded arrangements, she added, but as the conversation around speed-to-power has matured, most projects involve utilities, large loads, and XGS together from the start.
A wave of geothermal IPOs?
Fervo is acting as a test case for advanced geothermal on the public markets, but by most metrics its position is so far out ahead of other startups that it can be a tricky benchmark for companies like XGS, Diaz said.
Nonetheless, XGS is reportedly considering its own public offering. It appears to be at a far earlier stage than Fervo was when it started the process of going public. But there really isn’t a good baseline for the sector.
“With any new industry there are unknowns as to when [should] they go public,” Diaz explained. Fervo, for example, reported just $138,000 in revenue in 2025 before its IPO, which might feel a little early for many industries.
But drilling is extremely capital-intensive, and one of the things public markets provide is easier access to capital. The moment when geothermal stops being able to raise funds efficiently and with favorable terms from venture capital may be earlier than in other sectors, she added: “There’s some amount of figuring it out in real time — are public markets the right option for these companies?”
Other companies are undoubtedly looking at Fervo’s performance to evaluate their own options, she acknowledged. But as a whole, the industry needs to get more public with its pipelines and operating data.
“Broadly, we need to see disclosure of more technical results,” Diaz said. Investors want enough technical data to understand how much risk they’re taking on. It can be tricky for geothermal in particular, because getting a full understanding of the resource potential of a particular geography typically requires drilling, which is expensive. (That said, companies like Zanskar are deploying machine learning to reduce the need for that exploration.)
Fervo’s willingness to disclose preliminary results along the way played in its favor when it came to its IPO, Diaz said. To date, other companies — even those that may be eyeing a public offering — have seemed less prepared to do so, she added.
