What Oklo Actually Does (or Will Do)
Oklo is building small modular reactors — specifically a sodium-cooled fast reactor called the Aurora Powerhouse. Unlike traditional nuclear plants, which are single 1,000+ megawatt installations that take 15-20 years to permit and build, Oklo's Aurora design is modular: 15-50 MW units intended to be factory-fabricated and deployed at customer sites in months, not decades. The business model isn't selling reactors — it's selling power. Oklo plans to own and operate the reactors and sell electricity to customers under long-term power purchase agreements (PPAs), similar to how a solar developer finances, builds, and sells power from a solar farm.
The Aurora's fuel is HALEU — high-assay low-enriched uranium — which Oklo can produce in part by recycling spent nuclear fuel from existing U.S. reactors. The sodium cooling system runs at atmospheric pressure (unlike water-cooled reactors that require high-pressure containment), which Oklo argues makes the design inherently safer and simpler. The company was founded in 2013, went public via SPAC in 2024, and has never generated commercial revenue. No Aurora Powerhouse is currently operational.
Why AI Created This Opportunity
The AI buildout has a power problem. A standard server rack draws 5-10 kilowatts. An AI GPU rack running NVIDIA H100s draws 60-100 kilowatts — 10-20x more power density. As hyperscalers build out AI training and inference infrastructure at unprecedented scale, their power consumption is growing faster than the grid can absorb. Microsoft has committed to carbon neutrality. Google has pledged 24/7 carbon-free energy. Meta is targeting 100% renewable power. And every one of these companies is discovering that solar and wind don't solve the "always-on" problem — AI inference runs 24 hours a day, and the sun doesn't always shine.
Nuclear is the only zero-carbon energy source that provides baseload power — constant, 24/7 generation regardless of weather. The problem with conventional nuclear is the timeline and cost: $10-30 billion per plant, 15-20 years to build, decades of permitting. Small modular reactors promise to solve this with factory production and shorter timelines. If they work, they are the answer to AI's power problem. That "if they work" is what the entire Oklo valuation debate is about.
The Pipeline: GW-Scale Agreements
Oklo has accumulated a significant pipeline of signed agreements — though the word "signed" requires careful parsing. These are not revenue contracts. They are letters of intent, master power agreements, and development agreements that commit to partnership but are contingent on Oklo receiving its NRC license and successfully constructing commercial reactors. The key distinction: this pipeline represents demand intent from major customers, not contracted cash flow.
- Switch: 12 GW non-binding Master Power Agreement — the largest corporate power agreement ever announced. Switch operates hyperscale data centers across the U.S.; 12 GW of nuclear power to serve their facilities would be transformative if delivered.
- Meta: 1.2 GW agreement in Pike County, Ohio — advancing Oklo's plans for a 1.2 GW nuclear power campus to support Meta's data centers in the region.
- Equinix: 500 MW power purchase agreement — the global data center colocation leader signed a supply agreement as part of its diversified clean energy strategy.
- NVIDIA partnerships: Five collaborations, specifics not fully disclosed, tied to supporting AI data center infrastructure power needs.
The cumulative pipeline represents 13+ GW in agreement commitments. For reference, the entire U.S. nuclear fleet produces roughly 100 GW across approximately 90 reactors. Delivering a material fraction of this pipeline would make Oklo one of the largest electricity producers in the country. That's a 10-20 year project, not a 3-year one.
Sam Altman: The Signal Behind the Chairman
Sam Altman became Oklo's executive chairman when the company went public in 2024. Altman was an early investor before his role at OpenAI dominated his public profile. His involvement carries two signals worth taking seriously. First, the man running the company that is consuming more electrical power than most small countries has a direct personal interest in solving the nuclear power problem — the thesis isn't abstract for him. Second, Altman's credibility as a capital allocator and network operator has helped Oklo attract the hyperscaler interest reflected in the pipeline above. Whether OpenAI eventually becomes an Oklo power customer is the unspoken optionality embedded in the relationship.
The Regulatory Hurdle: NRC License
This is the single biggest risk in the entire thesis. In 2022, the U.S. Nuclear Regulatory Commission denied Oklo's first combined license application for its Aurora reactor — not because the design was unsafe, but because the application was incomplete. The NRC cited insufficient safety analysis documentation. Oklo has been in extensive pre-application dialogue with the NRC since and submitted a Principal Design Criteria (PDC) Topical Report that the NRC accepted for review under an accelerated timeline in September 2025. The NRC has signaled its commitment to modernizing licensing for advanced reactors, and the regulatory environment under the current administration has been constructive for nuclear development.
But here's the hard truth: Oklo cannot generate a single watt of commercial power until the NRC approves its combined license application. That process — even on an accelerated track — likely takes 2-3 more years from application submission. A realistic timeline for the first Aurora Powerhouse generating commercial power is 2028-2030. Every signed agreement, every partnership, every GW of pipeline is contingent on that license arriving and the construction going as planned. Regulatory risk is the tail risk that can zero the stock — it's a low probability but non-trivial event.
Q1 2026 Earnings — The Financials of a Pre-Revenue Company
Oklo reported Q1 2026 on May 12: EPS of -$0.19, beating the -$0.20 consensus by $0.01. There was no revenue. There will be no revenue until a commercial reactor is online. This is not a traditional earnings context — you're assessing cash burn rate, cash runway, and development milestones, not income statement performance.
- Cash & Investments: $2.209B — significant war chest, raised through the 2024 SPAC and subsequent capital markets activity
- Net Cash: $2.206B — essentially debt-free
- TTM Operating Cash Outflow: -$87.8M
- TTM CapEx: -$65.7M (increasing as development accelerates)
- TTM Free Cash Flow: -$153.5M
- Cash Runway (at current burn): ~14 years — not a near-term cash concern
The cash position matters: Oklo is not in danger of running out of money. At $153M in annual cash burn, they have 14+ years of runway before needing to raise capital. That runway is the strategic cushion that makes this a long-term speculation rather than a near-term bankruptcy risk.
Q2 2026 earnings are scheduled for August 7, 2026 — next week. Expect more of the same: no revenue, cash burn update, and development milestone commentary.
Today's 11.54% Move — Sector, Not Company
OKLO gained 11.54% on July 30, alongside NuScale Power (+11.92%) and X-Energy (+12.25%). The move was sector-driven — Holtec announced positive SMR development news, and a broader energy sector rally tied to rising oil prices from Middle East tensions lifted nuclear names along with traditional energy. There was no Oklo-specific catalyst. This is important context: when a pre-revenue nuclear stock moves 11% in a day without a company announcement, you're watching sentiment trading, not fundamental repricing. The stock can give that 11% back just as easily on a risk-off day.
Technical Milestones Worth Watching
Two near-term technical/regulatory milestones are meaningful:
- Groves Isotopes Test Reactor — First Criticality: Oklo's Groves test reactor at a DOE facility was targeting first criticality in July 2026. If achieved, this would be the first operational milestone for any Oklo reactor — not commercial, not Aurora, but proof that Oklo can actually make a fission reaction happen under its design parameters. First criticality is the difference between "company that has drawings" and "company that has made atoms split." Watch for this announcement.
- Aurora NRC COLA Submission: The combined license application for the first commercial Aurora Powerhouse is the pivotal regulatory document. Once submitted, the NRC review clock starts. Management has been in pre-application preparation for 18+ months.
Valuation — Honest Assessment
- Price (July 30 close): $41.09
- Shares outstanding: ~153M
- Market Cap: ~$6.3B
- Enterprise Value (ex-cash): ~$4.1B
- Revenue: $0
- Truist PT: $55 (Hold) — initiated July 2026
- First commercial power estimate: 2028-2030
There is no traditional valuation framework that prices OKLO today. You can't apply a PE multiple to zero earnings. You can't apply an EV/Revenue multiple to zero revenue. What you're doing instead is probability-weighting scenarios: if Oklo gets licensed and delivers even 2-3 GW of its pipeline by 2032, what is that power business worth? At typical utility valuations (5-7x EBITDA), a 2 GW base with 90% utilization at $70-80/MWh in power sales revenue approaches $8-10B in annual revenue — a substantially larger company than today's market cap implies.
The bear case is simpler: the license takes longer than expected, construction costs exceed projections, a competitor beats them to scale, or the AI data center buildout plateaus before Oklo has operational capacity. In any of these scenarios, the stock revisits the SPAC-era lows near $10-15.
Position Framework
- Setup: Pre-revenue advanced nuclear developer with GW-scale pipeline commitments, strong cash position, Sam Altman as chairman, and sector tailwind from AI power demand; NRC license is the single gating variable
- Entry zone: $32–$38 — any pullback into pre-sector-rally levels creates a cleaner entry point; today's 11% sector-driven move is not a signal to chase
- Add trigger: NRC accepts Oklo's combined license application for formal review (signals regulatory pathway is clear), OR Groves reactor achieves first criticality (proves the physics work)
- Trim zone: $60–$70 — toward the Truist $55 PT and above, if near-term catalysts (NRC filing, Groves milestone) drive sentiment re-rating
- Target: $80–$120 on a 3-5 year basis if first Aurora commercial operation is achieved and pipeline conversion begins; this is a scenario target, not a base case
- Invalidation: NRC rejects second license application, OR management significantly increases cash burn without corresponding development milestones, OR key pipeline agreements (Switch, Meta) are canceled rather than renegotiated
- Position sizing: This is a spec allocation — 1-3% of a portfolio maximum. The range of outcomes is extreme: 10x if it works, near-zero if it doesn't. Size accordingly.
- Time horizon: 3-7 years — the thesis resolves on a nuclear energy timeline, not a quarterly earnings timeline
Oklo is one of the most interesting pre-revenue speculations in the market right now — not because the stock is cheap, but because the problem it's trying to solve is genuinely important and the customer demand is clearly there. The pipeline of GW-scale agreements from Meta, Switch, and Equinix tells you the hyperscalers believe nuclear has a role in their power strategy. Whether Oklo specifically is the company that delivers on that demand is a question of execution, licensing, and time. At $41 and $6.3B in market cap, you're paying a meaningful premium for that optionality. At $32-38, that premium gets more defensible. The sector moved today — don't chase it. Wait for the pull, and wait for the NRC.