The opinions expressed here by Trellis expert contributors are their own, not those of Trellis or its editors.
In the natural life cycle of business and finance, companies get into trouble, and the first question asked is who has to pay for their products and whether those customers can stop buying. The quick second question is whether the business is fundamentally sound but poorly financed. Climate tech is now being sorted by both criteria.
In the third quarter, three companies each raised a billion dollars: a fusion developer, Commonwealth Fusion Systems; a fission developer, Valar Atomics; and Base Power, a company that leases home batteries and runs them as a power plant.
Firm power — fusion, fission, geothermal, storage — took in roughly $6.3 billion across 17 rounds, about 40 percent of all climate tech equity in the quarter, according to the third-quarter edition of the State of Climate Tech quarterly analysis.
Commonwealth, Valar and Base Power did not raise their money on the strength of their emissions math. They raised it because hyperscalers need electrons they cannot get from the electricity grid on the schedule their lenders require.
The aggregate numbers tell the same story: Capital is concentrating in fewer companies, nearly all of them selling to companies behind the artificial intelligence (AI) buildout. There is funding for fundamentally sound, promising climate tech that can demonstrate customer interest or demand. Equity funding rose 17 percent in the second quarter to $15.8 billion, while the number of deals fell to the fewest since 2021. In the first half, 10 deals took 42 percent of the money, and low-carbon data center developers alone took a third, up from 3 percent a year earlier. More money, fewer companies, one kind of buyer. This is not a recovery of climate tech. It is a resorting of it.
The litmus tests for funding decisions
The sorting mechanism described below predicts which of a sustainability leader’s priorities will have a vendor in three years and which will not. Every category still raising funds passes most of these five tests. Every category that has stalled fails at least two.
First test: a buyer who must pay. Data center developers are paying up to $150 per megawatt-hour, roughly a 50 percent premium over what a conventional new power contract commands, for power that is clean, firm and available quickly. That premium is the carbon price that climate tech never got from policy. It funds first-of-a-kind geothermal wells, microreactors and iron-air batteries that no carbon credit ever would.
Second test: revenue visibility before the technology is finished. The customer now shows up in the capitalization of the company. Nvidia signed a development deal with Valar before Sequoia priced it at $6 billion; National Grid’s venture arm anchored Joulent’s $1.75 billion. A Series B with an offtake behind it is project finance with a longer runway, and growth investors have noticed: Series B funding jumped to $3.6 billion in the quarter.
Third, policy that survived. Nuclear and geothermal kept their federal tax credits and their bipartisan sponsors through the 2025 rollbacks in the One Big Beautiful Bill Act. Offshore wind, green hydrogen, and much of the fuels complex did not. The watch item is that the Trump administration’s fiscal year 2027 budget request trims both the nuclear and geothermal offices at the Department of Energy. So, survival does not ensure future permanence.
Fourth, a visible exit. Fervo Energy’s $1.9 billion geothermal public market listing in May and X-energy’s $1 billion nuclear initial public offering reopened a window shut for three years; 17 climate companies went public in the first half. The glass is thinner than it looks. Both stocks are well off their highs — Fervo by roughly a third, X-energy by more than half — and third-quarter exits fell to 78 from 99. Public markets will pay for contracted backlog. They will not pay for a story.
Fifth, a story that can be told as security or resilience. Nuclear power offers energy independence. Grid flexibility is reliability. Earth observation, the satellite and weather-intelligence companies that raised heavily this year, sell to insurers and defense ministries. Critical minerals, from domestic mining and processing to battery recycling, are financed as onshoring initiatives. Each of these categories raised in 2026 without saying the word carbon — the one framing that does not raise money this year, and the one the field was named for.
What doesn’t pass muster
Now run the failing categories through the same tests. Carbon removal fails the first, second, and fourth: no buyer who must pay for these solutions, no revenue before the tons are delivered, no exit. Equity funding for this category fell 61 percent in the first half, the weakest stretch since 2020, after Microsoft paused
Read More
