Oklo Inc. (OKLO) | The Buildout — AI Infrastructure
The Verdict
Oklo is developing small fast fission reactors it calls Aurora powerhouses, and plans to sell the electricity and heat they produce directly to customers under long-term, build-own-operate contracts. The firm, around-the-clock output that makes the design useful to a utility or an industrial site is the same output a data center needs, so the tie to AI is who buys the power rather than a product designed only for AI. The company describes itself as one platform across three lines: power, fuel and isotopes, with power as the anchor, fuel as the enabler and isotopes broadening the customer set. Management also says it wants third-party capital to fund a larger share of each powerhouse, a structure it says could lower the direct capital Oklo invests per deployed megawatt.
| Market Cap | — |
| Revenue (TTM) | $1M |
| EBITDA Margin (TTM) | -18091.7% |
| Net Cash | $1.6B |
| Earnings Beats | 2 of 7 |
| P/E (TTM) | — |
| EV/EBITDA (TTM) | — |
What We Like
- Groves, the privately sited isotope reactor, reached first criticality in early August 2026, after 229 days of substantial construction and less than a year from groundbreaking. Management calls it the fastest greenfield-to-criticality transition for a full-scale privately funded and privately sited reactor that it is aware of.
- Named demand includes a 1.2 GW Meta power campus in Pike County, Ohio under a prepayment agreement dated January 5, 2026, and a 12 GW Master Power Agreement with Switch signed in December 2024 that the 10-K calls one of the largest corporate power purchase agreements in history.
- Fuel is being sourced through several paths: a DOE award of five metric tons of HALEU from EBR-II recovered material, a Centrus letter of intent covering up to five Aurora powerhouses with deliveries expected from 2029, and a DOE surplus plutonium tranche of 20 tons that management says is equivalent to 160-200 tons of HALEU.
- The balance sheet carried $3.0 billion of cash and marketable securities at Q2 end, split between $1.6 billion of cash and $1.4 billion of marketable securities, and $1.9 billion of capital generated from ATM programs in 2026.
- Aurora-INL cleared a licensing step when DOE approved the Preliminary Documented Safety Analysis, and Kiewit is the selected lead constructor there and signed an MOU for initial-phase execution planning in Ohio.
What We’re Watching
- Aurora-INL full project cost has not been disclosed. Management says it is still narrowing the total with Kiewit and will provide it through the rest of this year.
- First isotope revenue slipped from prior 2026 language to the first part of next year, and the source is now described as more likely the Idaho lab than Groves.
- Centrus HALEU deliveries are expected to begin in 2029, after the 2028 Aurora-INL start-up target, and the letter of intent is not yet a definitive agreement.
- Part 57 fleet-based licensing went silent in Q2 2026 after a detailed Q1 discussion, and the Eielson site characterization update also went silent.
Execution credibility strengthened this quarter: Groves moved from a target to a completed asset, and a DOE safety analysis approval corroborated it. Demand evidence stayed strong but is still mostly non-binding, and the nearest revenue line moved from 2026 to early 2027. Fuel visibility improved on paper, though the Centrus letter of intent is not definitive and its delivery date sits after the Aurora-INL start-up target. The case now depends on whether cost disclosure and fuel contracting arrive fast enough to hold the 2028 date.
Earnings
Oklo reported $1.2 million of revenue in the quarter and a 40.4% gross margin, with EBITDA of negative $72.9 million and a net loss of $48.5 million. The standout operating event was Groves reaching first criticality in early August 2026, and on the same call management raised its 2026 cash-spend guidance for both operating activities and property, plant and equipment.
| Metric | Q2 FY2026 | Q1 FY2026 | Q2 FY2025 | YoY |
|---|---|---|---|---|
| Revenue | $1M | $0M | $0M | — |
| Gross margin | 40.4% | — | — | — |
| EBITDA | −$73M | −$51M | −$28M | +161.3% |
| EPS | $-0.28 | $-0.19 | $-0.18 | +56.1% |
| Cash and marketable securities | $3.0B | $2.5B | n/a | — |
We completed substantial construction in 229 days and reached first criticality in less than a year after groundbreaking.— Jacob DeWitte, Chief Executive Officer, 2026-08-07
Management tone: Management moved from promising execution to reporting it. The prior call framed the story as increasingly about execution; the Q2 2026 call led with a completed milestone and added first-of-its-kind language around Groves. On the same call the CFO raised both 2026 cash-spend guides with a mechanistic explanation and pushed first isotope revenue to the first part of next year. Management did not update Part 57 or the previously mentioned Centrus deconversion joint venture, and it deferred the Aurora-INL all-in cost when asked.
Management Guidance
On the Q2 2026 call, CFO Craig Bealmear raised FY2026 cash used in operating activities to $120 million-$150 million from a prior $80 million-$100 million, and raised cash used for property, plant and equipment to $400 million-$500 million from a prior $350 million-$450 million. Management attributed the increase to first-of-a-kind project costs at Aurora-INL that were expensed, grid-interconnection build-out, and acceleration of early-stage deployment costs on other projects, and said it did not reflect materially higher general corporate burn. Aurora-INL start-up remains targeted for 2028. Full Aurora-INL project cost was not provided; management said it is still narrowing the number with Kiewit and will disclose it through the rest of this year.
Trajectory
The reported numbers are still dominated by spending. Trailing revenue is $1.2 million against EBITDA of negative $217.1 million, a negative 18,091.7% EBITDA margin, and trailing free cash flow of negative $275.9 million. Through the first half of 2026 the company reported a net loss of $81.6 million, an operating loss of $124.2 million and $65.5 million of cash used in operating activities, partly offset by $44.5 million of net interest and dividend income, with capital spend of $126.9 million. Both 2026 cash-use guides were raised, and management ties the increase to project work rather than general corporate purposes. The direction of travel is a widening loss as spending is pulled forward to protect the 2028 target.
The Model
The model projects FY+1 revenue of $9.6 million and EBITDA of negative $289 million, a negative 3,010.85% margin, and FY+2 revenue of $22.35 million and EBITDA of negative $417 million, a negative 1,865.35% margin. The near-term revenue figure is small; the only near-term revenue the company itself points to is first isotope sales in the first part of next year. The larger FY+2 line assumes that base builds, and the model carries a widening EBITDA loss in both years, consistent with guidance that points to higher 2026 cash spend as work is pulled forward.
| Metric | FY2025 | Next FY (E) | Following FY (E) |
|---|---|---|---|
| Revenue | $0M | $10M | $22M |
| YoY Growth | — | — | +132.8% |
| EBITDA | −$139M | −$289M | −$417M |
| EBITDA Margin | — | -3010.8% | -1865.3% |
Projections are the median of 4 independent model runs.
On the Q2 2026 call, CFO Craig Bealmear raised FY2026 cash used in operating activities to $120 million-$150 million from a prior $80 million-$100 million, and raised cash used for property, plant and equipment to $400 million-$500 million from a prior $350 million-$450 million. Management attributed the increase to first-of-a-kind project costs at Aurora-INL that were expensed, grid-interconnection build-out, and acceleration of early-stage deployment costs on other projects, and said it did not reflect materially higher general corporate burn. Aurora-INL start-up remains targeted for 2028. Full Aurora-INL project cost was not provided; management said it is still narrowing the number with Kiewit and will disclose it through the rest of this year.
What Could Go Right — and Wrong
- The Centrus letter of intent converts into a definitive HALEU agreement with disclosed volume, pricing and prepayment terms.
- DOE finalizes a plutonium allocation to Oklo with a defined tonnage; management says the 20-ton tranche supports a couple of gigawatts depending on allocations.
- Aurora-INL project cost comes in at a level management can fund, and the 2028 start-up target holds through documented safety analysis approval, readiness review and start-up authorization.
- First isotope revenue lands in the first part of 2027 as guided, giving the company a commercial product before power revenue begins.
- Third-party capital is committed at the asset or project level, the structure management says could reduce the direct capital Oklo must invest per deployed megawatt.
- Fuel pathways slip: Centrus deliveries begin in 2029 and the letter of intent is not definitive, so the 2028 Aurora-INL target depends on EBR-II material, A3F output, plutonium or inventory bridging the gap.
- Aurora-INL cost or schedule slips; management has not yet disclosed a full project cost and says it is still narrowing the number with Kiewit.
- PJM interconnection turnaround for the Ohio campus extends, which management itself calls the watch point.
- Part 57 fleet-based licensing fails to finalize, and the Tennessee recycling facility's NRC application readiness work stretches out.
- First isotope revenue slips again beyond the first part of next year, the timing guided on the Q2 2026 call.
Looking Ahead
Over the next 12 months the visible checkpoints are Aurora-INL project cost, promised through the rest of this year, and first isotope revenue, guided to the first part of 2027. Behind those sit the Aurora-INL licensing steps that keep the 2028 start-up in view: documented safety analysis approval, readiness review and start-up authorization. Also running are A3F installation and start-up in 2027, conversion of the Centrus letter of intent into a definitive agreement, and the DOE plutonium allocation, whose timing management says is not under its control.
- Rest of 2026Aurora-INL cost guidance — Management promised a narrowed project cost with Kiewit.
- Early 2027First isotope revenue — Guided to the first part of next year, likely from the Idaho lab.
- 2027A3F installation, start-up — Fuel fabrication equipment installed and started up.
- 2028Aurora-INL start-up — Target holds; hinges on DSA, readiness review, authorization.
- 2029Centrus HALEU deliveries — Up to five Aurora powerhouses; letter of intent not definitive.
Financials
Annual Summary
| Metric | FY2024 | FY2025 | TTM | YoY |
|---|---|---|---|---|
| Revenue | $0M | $0M | $1M | — |
| Gross Margin | — | — | 40.4% | — |
| EBITDA | −$52M | −$139M | −$217M | -164.4% |
| EBITDA Margin | — | — | -18091.7% | — |
| Net Income | −$72M | −$106M | −$153M | -47.5% |
| Free Cash Flow | −$39M | −$115M | −$276M | — |
| Net Cash | — | — | — | — |
Key Ratios (Trailing)
- P/E TTM—
- EV/EBITDA TTM—
- EV/Revenue TTM—
- Price/FCF TTM—
- Gross Margin (TTM)40.4%
- EBITDA Margin (TTM)-18091.7%
- Net Margin (TTM)-12725.0%
- ROIC-10.5%
- SBC / Revenue3541.7%
The Company
Oklo is developing next-generation fast fission power plants called Aurora powerhouses, meant to produce electricity and heat it sells directly to customers. The 10-K describes the plan as selling that power and heat under a build-own-operate model, where Oklo keeps the asset and sells the output under long-term commercial arrangements, while also commercializing nuclear fuel recycling. The company reports a single reportable segment, with the CEO as the chief operating decision maker, and names three product lines: the Aurora powerhouse, nuclear fuel recycling and radioisotopes. Management describes these as one platform, with power as the anchor, fuel as the enabler and isotopes expanding the platform's value.
Oklo is bringing more of the chain in-house rather than buying a finished product. It holds a DOE site use permit at Idaho National Laboratory, where the first Aurora powerhouse and the Aurora Fuel Fabrication Facility sit, and it is designing a used-fuel recycling facility in Oak Ridge, Tennessee as the first phase of an Advanced Fuel Center that the 10-Q describes as a roadmap of up to $1.68 billion of investment and more than 800 jobs. It acquired ARMEC, an Oak Ridge manufacturer already doing engineering and procurement work for Aurora-INL, and Creative Engineers, which brought roughly 30 years of sodium and alkali-metal experience. The 10-K warns the company relies on a limited number of suppliers, some designed for first-of-a-kind or sole use, and names Siemens Energy, Amentum, Kiewit and Centrus as supplier risks. Principal executive offices are in Santa Clara, California, under a sublease.
Business Segments
Competitive Landscape
Oklo competes in advanced nuclear. The source material does not describe a sole-source position for the company; it lists a named set of rival developers across small modular and fast reactor designs. Centrus appears in three roles at once: HALEU supplier under the letter of intent, competitor, and previously counterparty to a deconversion joint venture that went silent. One named competitor, Nano Nuclear, has publicly argued on its own call that a fast reactor would need a very enriched central core of fuel and that HALEU is not available.
- Centrus EnergyAppears as a HALEU supplier under the letter of intent and as a competitor: the 10-K says a small but growing supply is being produced domestically and made available at high prices by Centrus and others.
- TerraPowerNamed in the competitive set as a sodium fast reactor developer. Named in filings; not discussed.
- X-EnergyNamed as an advanced reactor developer. Neighbor data notes Centrus holds a definitive HALEU offtake with X-energy.
- Nano NuclearNamed as a microreactor competitor. On its own call, that company argued a fast reactor would need a very enriched central core of fuel and that HALEU is not available.
- Named in the competitive set with the BWRX-300. Named in filings; not discussed.
Supply Chain
Oklo sits at the center of its own chain: it designs the reactors, has selected a lead constructor, and is bringing fuel sourcing, fabrication and recycling in-house rather than buying a finished product. Among the counterparties whose own calls were reviewed, only Centrus names Oklo.
More on OKLO: Earnings recap