Small Modular Reactor Companies You Can Buy and the Risks Involved

Search for “SMR stocks” and you might expect a crowded, mature sector. In reality, only a handful of small modular reactor companies trade on a public exchange, and their prices can halve in months. NuScale Power was valued at roughly US$5.3 billion in March 2026. By the 9 October 2026 close, that figure had fallen to about US$2.95 billion.

Power demand from artificial intelligence and data centres has pushed these reactors into mainstream investor conversation. Yet most developers earn no revenue yet and sit years away from producing commercial electricity.

That leaves you with a real trade-off. Getting in early could give you exposure to a long-term structural theme. Getting it wrong could tie up your capital in volatile names that move on stories rather than earnings.

This guide maps which developers you can actually buy, how far each has progressed on licensing and construction, and which risks deserve the most weight before you commit money.

This is general information only, not personal financial advice.

What are small modular reactors, and which companies can you actually buy?

A small modular reactor (SMR) is a nuclear reactor built mostly from factory-made modules that are shipped to site and assembled. The aim is faster construction and a lower upfront cost per unit than a conventional plant. The common benchmark is output below 300 MW, compared with roughly 1,000-1,600 MW for a large reactor.

If the modular concept is new to you, the factory-built reactor design behind these projects explains why developers expect lower upfront costs per unit, and why the label matters less than the engineering approach you are buying into.

That definition is a useful starting point. It is also looser than it sounds.

Why “small” does not always mean under 300 MW

Several leading designs break the 300 MW line at plant level. Rolls-Royce SMR is rated at 470 MWe, TerraPower’s Natrium at 345 MWe nominal, and a six-module NuScale plant at 462 MWe, even though each NuScale module is just 77 MWe. In practice, “small” describes the modular, factory-built approach more than a strict size ceiling, so judge each company on its design rather than the label.

The investable universe is even narrower than the definition. These are the listed pure plays:

Figures as of 9 October 2026.

Many of the names you will see in headlines cannot be bought directly on the sharemarket:

• TerraPower (private, backed by Bill Gates)
• Rolls-Royce SMR (described as private; current ownership details not confirmed)
• Kairos Power (private)
• X-energy is reported to have listed in April 2026, raising US$1 billion, according to Reuters (23 June 2026)

The gap between three or four listed names and a much longer list of private developers tells you something important. Direct SMR exposure is narrow, so most of your decisions come down to a few volatile stocks or to indirect routes.

SMR developer profiles: where each company stands on licensing and construction

With the map drawn, the next question is how far each developer has actually travelled. Compare them on three things: design, regulatory position and physical progress on the ground.

NuScale

NuScale is reported to hold the only full design certification or approval from the US Nuclear Regulatory Commission (NRC) for an SMR. Its uprated 77 MWe US460 module received Standard Design Approval in May 2025, and the design relies on passive safety systems that need no external power to cool the reactor. A Romanian project at Doicești with RoPower has been reported as reaching final investment decision.

A reminder from NuScale’s own history The company’s original Carbon Free Power Project was cancelled. Regulatory approval did not guarantee a built plant, and you should not assume it will next time either.

Oklo

Backed by Sam Altman, Oklo is developing the Aurora microreactor, cited at 15 MW, with some sources quoting a 15-75 MWe range. The company reportedly broke ground at Idaho National Laboratory in September 2025. Its licence application is in progress, though the exact docket stage could not be confirmed.

TerraPower

Natrium is a sodium-cooled fast reactor rated at 345 MWe nominal and 500 MWe peak. The NRC completed its environmental review in October 2025 and its final safety evaluation in December 2025, with a construction permit decision expected in the first half of 2026. That outcome is not confirmed in available research, so check NRC records before relying on it. Yale reported in April 2026 that non-nuclear construction had begun at Kemmerer, Wyoming.

Rolls-Royce SMR and X-energy

Rolls-Royce SMR’s 470 MWe pressurised water reactor is the largest SMR design. It has reportedly made UK regulatory progress, with Wylfa as its preferred site and a reported 599 million pound loan facility from the UK National Wealth Fund.

X-energy plans four Xe-100 units (80 MWe each, 320 MWe total) at Dow’s Seadrift site in Texas. Reports differ on timing, so expect an NRC construction permit around late 2026 to early 2027.

Other developers

Terrestrial Energy remains far smaller than its listed peers. Moltex Energy and Global First Power are private, and their current corporate status could not be confirmed, so no detailed claims are made here.

The spread is wide, and some “progress” is still early site work. Because licensing and first-concrete dates keep moving, read any company’s timeline as a range and treat slippage as a normal, recurring risk.

The NRC’s first commercial SMR construction permit went to TVA for the GE Vernova Hitachi BWRX-300, a reminder that the most advanced US project is not one of the listed pure plays.

Hyperscaler deals: what Amazon, Google and others have actually committed to

Regulatory milestones are one driver. The other is demand, and few signals excite the market more than a technology giant signing on.

The demand gap The International Energy Agency (IEA) reports data-centre electricity demand grew 17% in 2025, against 3% growth in overall demand. It projects data-centre power use roughly doubling by 2030.

Look past the headlines, though, and the agreements contain dates that sit mostly in the 2030s:

1. Google and Kairos Power (14 October 2024): up to 500 MW of capacity by 2035, starting with a first 50 MWe unit around 2030, followed by 75 MWe modules.
2. Amazon and X-energy (October 2024): Amazon reportedly led a US$500 million financing round and partnered with Energy Northwest on up to 12 Xe-100 units in Washington state. The Cascade Advanced Energy Facility near Richland is reported to start with four units totalling 320 MWe, with construction expected to begin by the end of the decade. These details were not independently confirmed.
3. Equinix and Meta with Oklo: agreements have been reported, but capacity figures could not be verified.

Current progress on all of these should be confirmed through company filings, since status checks were limited.

Hyperscalers are also paying for nuclear power uprates at operating reactors, with roughly 1.1 GW funded under 20-year contracts, a faster route to new capacity than waiting for 2030s SMR deliveries.

A May 2026 research note put hyperscaler willingness to pay at a reported US$107-130/MWh for firm clean power. That tells you buyers will pay a premium for reliable low-carbon electricity. The 2030s delivery dates, however, mean these deals support the long-term case without proving near-term earnings, so ask how much of any share price already assumes they will be delivered on time.

Risks, valuation and how to gain exposure: pure plays versus indirect routes

Distant deliveries and moving timelines show up directly in share prices. Oklo’s market capitalisation fell from about US$12.9 billion in March 2026 to about US$6.45 billion by 9 October, while NuScale dropped from US$5.3 billion to US$2.95 billion.

A halving in roughly seven months tells you these stocks trade on sentiment and milestones. Size any position as money you can afford to see fall sharply.

The nuclear equity selloff has hit developers with strong technical progress as hard as those with little, which is why milestones alone have not protected valuations in recent months.

The bull case versus the cautious case

Supporters point to premium-priced hyperscaler contracts and government financing support, such as the reported UK National Wealth Fund facility. Sceptics point to pre-revenue companies, unproven business models and earlier gains that were reported at more than 400% for NuScale and nearly 600% for Oklo since early 2024. Both views rest on real evidence, which is why position size matters so much.

Key risks to price in

• Funding and dilution: with no revenue, developers rely on equity raises, grants and partners, and new shares shrink your slice of the company.
• First-of-a-kind cost overruns: first units have historically exceeded budgets.
• Regulatory delay: permit dates shift, as X-energy’s differing targets show.
• HALEU fuel supply: limited enrichment capacity could slow fast-spectrum designs such as Natrium.

Two terms need plain definitions. A first-of-a-kind (FOAK) unit is the first commercial build of a new design, so it carries higher costs and more unknowns than later copies. Yale’s April 2026 analysis reportedly found these units remain capital-intensive and dependent on government support.

HALEU stands for high-assay low-enriched uranium, which is uranium enriched to a higher level than standard reactor fuel. Supply is limited and geographically concentrated, and quantified cost projections were not found.

Pure plays or indirect exposure?

If you want pure-play exposure, many investors keep it to a small portion of the portfolio and pair it with diversified holdings. If a large drawdown would keep you up at night, indirect routes give you the theme with more cushioning when individual projects slip.

Past performance does not guarantee future results. Forward-looking statements are speculative and subject to change.

Weighing the SMR opportunity without chasing the headlines

The buyable list is short, milestones drive valuations, and hyperscaler deals strengthen the long-term case while delivery sits in the 2030s. Before you buy, check three things:

1. Current regulatory status, using company announcements and NRC filings.
2. Funding runway and dilution risk, including recent or planned capital raises.
3. Position size, measured against your total portfolio and your tolerance for sharp falls.

All market figures here are as of 9 October 2026 and will have moved since. Verify prices and project statuses with primary sources before acting.

This article is for informational purposes only and should not be considered financial advice. Investors should conduct their own research and consult with financial professionals before making investment decisions.