Nuclear Startup Investment Reaches $4.5 Billion in 2026, Tracking to Beat Annual Record
Capital flowing into 81 fission and fusion companies is on pace to surpass 2025's $6.2 billion full-year total, driven by AI power demand projections.
Global investment in nuclear fission and fusion startups topped $4.5 billion across 81 companies through mid-2026, Axios reported, with oilprice.com covering the figures on Friday (2026-07-31). At that run-rate, the sector is on course to shatter 2025's record of $6.2 billion — itself spread across 93 companies.6
The 2025 comparison is instructive. Last year's record came in a sector already attracting significant capital, and 2026 is doing more with fewer companies — 81 this year versus 93 then — suggesting larger average round sizes even if the data does not break them out explicitly.6
The driver is electricity demand. U.S. data-center power consumption is projected to more than triple over the next decade, from 34.7 gigawatts in 2024 to 106 GW by 2035, according to the IEA — a trajectory that concentrates load in geographic clusters and at hours when solar generation is absent. Nuclear's high capacity factor makes it attractive precisely because it does not require backup storage to perform.3
The EIA's Annual Energy Outlook 2026 adds longer-range context. Server consumption in commercial buildings alone could reach between 446 billion and 818 billion kilowatthours by 2050, the agency projected. Servers already represented an estimated 7% of commercial-sector electricity consumption in 2025, and that share climbs to between 22% and 33% of total commercial building load by mid-century across EIA's scenarios.1
That is the arithmetic venture capital is working from. Washington has separately stated a target of quadrupling U.S. nuclear capacity by 2050, Forbes reported on May 28 (2026-05-28), though no specific federal funding commitment has been attached to that goal in available reporting.3
Goldman Sachs has included small modular reactors in its global reactor tracker, its latest "Nuclear Nuggets" publication shows — a sign that institutional finance is treating SMR timelines as live investment variables rather than speculative placeholders. But tracker inclusion does not equal commissioning. Advanced reactor projects still face multi-year permitting and construction processes, and the gap between capital raised and megawatts delivered has historically been wide.2
China adds competitive urgency to the backdrop. Its Experimental Advanced Superconducting Tokamak fusion reactor was reported to be on track to achieve ignition in 2027, oilprice.com reported on June 12 (2026-06-12), a milestone that would precede any commercial fusion project in the United States if it holds. U.S. private fusion firms are racing to close that symbolic gap while managing investor expectations on commercialization timelines that stretch well past 2030.5
Public market signals on Friday (2026-08-01) cut against the private funding enthusiasm. The URA uranium ETF fell 2.01% that session. Uranium spot prices and SMR investment rounds track different things — URA reflects near-term fuel demand rather than equity stakes in advanced reactor startups — but the divergence illustrates that equity investors are not uniformly convinced nuclear's lead times will produce returns within their hold periods.6
For U.S. grid context: EIA projects solar generation in ERCOT will reach 78 billion kilowatthours in 2026, outpacing coal at 60 BkWh in that grid. The crossover shows solar's deployment speed but also underlines why firmer capacity sources remain a priority for grid operators and corporate buyers who need around-the-clock coverage that intermittent generation cannot provide alone.1
The IEA noted in its World Energy Investment 2026 report that the West Asia conflict and Hormuz disruptions have pushed nations toward domestic energy investment including nuclear, layering a geopolitical motivation onto the commercial one. That framing has reinforced the policy case for nuclear even as reactor lead times remain unchanged.4
By December, the full-year tally will either confirm a new record or reveal a second-half slowdown in deal flow. More revealing will be the breakdown between near-term fission projects with active regulatory filings and longer-dated fusion concepts whose commercial power delivery extends past 2040. The Axios report and oilprice.com coverage do not specify that split — and it is precisely that detail which matters most to grid planners sizing capacity against the IEA's 106-GW data center load projection for 2035.6,3