Commonwealth Fusion Systems
Pre-revenue, capital-intensive deep-tech developer, vertically integrated across the fusion value chain (HTS-magnet manufacturing → tokamak design/build → power-plant operation). Future monetization: electricity sales via PPAs, technology/magnet licensing, and HTS-magnet supply to third parties. No commercial revenue today.
CFS has never officially disclosed a post-money valuation for any round — its press releases (cfs.energy, PRNewswire) confirm raise size only. So this trail is built from credibly reported THIRD-PARTY estimates, not company figures; both points are tagged kind='secondary' to reflect that no primary round mark exists. Trail of actual rounds (all with UNDISCLOSED valuations, hence omitted as valuation points per the honesty gate): Series A ~$215M cumulative (2018–2020, Eni/Breakthrough/Khosla/Temasek/Equinor); Series B $1.8B (Dec 2021, lead Tiger Global) — the ONLY round with a credible third-party post-money estimate, CB Insights' $3.2–5.0B range (midpoint $4.0B shown; treat as a wide band, not a precise mark); Series B2 $863M (Aug 2025, NVIDIA/NVentures, Google, Breakthrough, Counterpoint Global/Morgan Stanley, Mitsui consortium) — NO valuation disclosed or reliably estimated, so OMITTED; ~$3.85B round (May 2026, Singapore-linked + existing backers; total funding ~$6.85B) — NO valuation disclosed, OMITTED. The ~$8B figure is a current aggregator estimate (appears in compworth/PitchBook-style profiles), NOT tied to the May-2026 round price; I could not confirm an $8B round post-money from any primary or reputable-press source — the user's '~$8B 2026 secondary' lead matches an aggregator/secondary-market estimate, not a disclosed primary mark. Nasdaq Private Market showed a secondary per-share NPM estimate of $195.23 with a last trade of $234 as of 2026-06-12, but NPM does not publish the implied total valuation, so no clean secondary cap is includable. Confidence is LOW: (a) zero company-disclosed valuations across the entire history; (b) the one historical estimate is a wide CB Insights range; (c) the current $8B is an aggregator estimate not traceable to a primary round price. Could not access Forge, PitchBook, or CB Insights detail pages (paywall/403). No SEC filings — CFS is private, no S-4/SPAC. Gaps: a precise 2021 post-money, and any disclosed post-money for the 2025 B2 or 2026 mega-round, were never published.
Earnings, margins, COGS & capex
Pre-revenue, equity-funded deep-tech developer. No electricity sales and no audited public financials. Economics are entirely forward-looking: value rests on the SPARC net-energy demonstration (target 2027) and the subsequent ARC ~400 MWe plant (early 2030s). Funded by ~$6.85B of cumulative equity (Series A through the $3.85B May-2026 round); cash burn driven by SPARC construction, the HTS-magnet factory, and R&D headcount (~1,000+ employees, 2025).
Revenue trend
Margins
Burn rising with SPARC build
Negative through commercialization
COGS structure
No cost of goods sold — no product shipped. Cost base is R&D + SPARC construction + HTS-magnet manufacturing scale-up + facilities (Devens, MA) and ~1,000+ staff. Not itemized publicly.
Capex
Extreme and front-loaded: multi-$B cumulative on the SPARC tokamak build and magnet factory; the ARC power plant will require billions more of unfunded capex. Capex intensity is the defining financial feature.
Latest earnings
n/a
No financial guidance. Operational milestones: SPARC first plasma and net-energy gain targeted 2027; ARC online early 2030s.
- Total raised (since 2018)
- ~$6.85B
- Latest round
- $3.85B (May 2026)
- Est. valuation
- ~$8B (2026; exact post-money undisclosed)
- SPARC net-energy target
- 2027 (Q>1; expected Q≈2–11)
- ARC capacity / online
- ~400 MWe / early 2030s
- Headcount
- ~1,000+ (2025)
Growth drivers
- SPARC net-energy demonstration (Q>1) — the gating scientific catalyst, target 2027.
- ARC ~400 MWe commercial plant in Virginia — first revenue (electricity) in the early 2030s if SPARC succeeds.
- PPA pipeline (Google 200 MW, Eni $1B) converting to contracted cash flows on delivery.
- Independent HTS-magnet sales/licensing as a potential near-term revenue line.
- Continued strategic + sovereign capital access (NVIDIA, Google, Temasek, Mitsui, Singapore).
Bull & bear
CFS is the most-capitalized, most-de-risked private fusion bet: a real tokamak under construction with a credible physics basis (MIT ARC papers), a Nobel-grounded HTS-magnet edge already validated by hardware tests and the DOE, ~$6.85B raised, and blue-chip offtake (Google 200 MW PPA, Eni $1B). If SPARC hits Q>1 in 2027, it is first to a commercially relevant net-energy machine and the option value on grid-scale fusion in the early 2030s is enormous.
- Most-funded private fusion company globally — ~$6.85B raised across A/A2/B/B2 and a $3.85B May-2026 round; capital depth is itself a moat in a field where most rivals are sub-$2B (Sacra, 2026).
- Hardware, not slideware: 20-tesla HTS (REBCO) magnet demonstrated in 2021; first of 18 SPARC toroidal-field magnets completed Jan 2026; SPARC building substantially constructed in Devens, MA (CFS/Wikipedia, 2026).
- Physics pedigree: direct MIT PSFC spinout; the high-field-tokamak path shrinks reactor size (smaller = cheaper) and rests on peer-reviewed ARC design work, the most conventional/least-exotic confinement approach among well-funded startups.
- Commercial validation already booked: Google signed a first-of-its-kind 200 MW corporate PPA on ARC and joined the B2 round; Eni committed $1B for power from CFS's first reactor (Sept 2025) — demand-side proof rare at this stage.
- Government tailwind: selected for the DOE Milestone-Based Fusion Development Program (eligible for up to ~$15M, most of 8 companies), with DOE-validated magnet milestones and an Energy-Secretary SPARC visit (CFS, 2026).
- Strategic-investor flywheel: NVIDIA (NVentures), Google, Breakthrough Energy, Temasek, Mitsui consortium, and Singapore-aligned investors in 2026 — names that bring data-center offtake demand and supply-chain reach, not just cash.
- Optionality on a second product line: the HTS-magnet capability is independently valuable (MRI, grid, accelerators, other fusion programs) even if SPARC slips — a partial floor under the equity.
Fusion has never delivered sustained, commercially relevant net electricity, and CFS is pre-revenue with a timeline that has already slipped (first plasma 2025→2027). Even a 2027 scientific Q>1 is years and many more billions short of a grid-connected, economic power plant; the ARC early-2030s date is aspirational. This is a binary, financing-gated, decade-plus call where dilution, technical surprises (disruptions, tritium fuel cycle, materials/neutron damage), and cheaper competing clean-firm power (fission SMRs, geothermal, solar+storage) can all impair or zero the equity.
- Pre-revenue with no commercial product — every figure that matters (electricity cost, plant capex, uptime) is a projection; CFS has never sold a kWh and won't before the 2030s on its own timeline.
- Timeline slippage is already visible: SPARC first plasma moved from 2025 to 2027, and scientific Q>1 ≠ a commercial plant — ARC at ~400 MWe in the 'early 2030s' assumes SPARC succeeds and the next, much larger build goes to plan.
- Brutal capital intensity: ~$6.85B raised and the expensive part (ARC) is still ahead — multiple further multi-$B rounds are needed, each diluting early holders; a fusion 'winter' or rate-driven capital pullback could strand the program.
- Unsolved engineering beyond Q>1: tritium breeding/fuel cycle, first-wall/divertor neutron damage and materials lifetime, disruption mitigation at 20 T, and magnet serviceability are hard problems no one has solved at power-plant scale.
- Intense, well-funded competition for the same prize: Helion ($15.5B val, different physics, Microsoft 2028 deal), TAE ($6B via TMTG merger), Pacific Fusion ($1B+ Series A), Zap/Type One/Tokamak Energy, plus public ITER — and fission SMRs (Oklo, NuScale, X-energy) chase the same clean-firm/data-center demand sooner.
- Offtake agreements are contingent and small relative to need: a 200 MW Google PPA and Eni's $1B are conditional on a plant that doesn't exist; they de-risk demand, not delivery.
- Illiquid, opaque private equity: no public price, no audited financials, valuation set by primary rounds/secondaries (~$8B est.); a down-round or extended private limbo is a real path. Conviction is LOW by design.
What it is worth
Private-round / scenario-based option valuation (no DCF — pre-revenue). Anchored to the implied ~$8B 2026 mark from the $3.85B round and secondaries, then probability-weighted across SPARC/ARC outcomes and benchmarked vs. fusion peers (Helion ~$15.5B, TAE ~$6B).
$2–4B or a down-round / extended private limbo if SPARC slips materially past 2027 or underdelivers on Q, capital tightens, and a cheaper clean-firm alternative (SMRs) captures the demand — early holders heavily diluted or impaired.
~$8–10B
roughly the current implied mark, holding as the $3.85B war-chest funds SPARC toward a 2027 result; modest re-rate on milestone progress, offset by dilution and field-wide timeline skepticism.
$15–25B+
if SPARC demonstrates a strong Q>1 in 2027 on schedule, de-risking the path to ARC and re-rating CFS toward/above Helion; the offtake pipeline (Google, Eni) converts to bankable contracts and the IPO window opens.
No public price; valuation is set by primary rounds and secondaries and is highly sensitive to the SPARC 2027 result. As an unprofitable, pre-revenue, financing-gated fusion bet, this is a venture-style option, not a cash-flow asset — conviction LOW. Exact May-2026 post-money is not disclosed; ~$8B is an estimate.
SWOT
Strengths
- Most-funded private fusion company globally (~$6.85B raised) — capital runway no rival matches.
- Validated HTS (REBCO) 20-tesla magnet technology — the core enabling edge, demonstrated in hardware and DOE-validated.
- Direct MIT PSFC pedigree and peer-reviewed high-field-tokamak (ARC) physics basis — the least-exotic well-funded approach.
- Real construction progress: SPARC building + first TF magnet (Jan 2026), not just simulations.
- Blue-chip demand and investor roster — Google PPA, Eni $1B, NVIDIA, Breakthrough Energy, Temasek, Mitsui, Singapore-aligned capital.
Weaknesses
- Zero revenue; no commercial product; deeply negative cash flow funded entirely by equity.
- Timeline has already slipped (first plasma 2025→2027); scientific net energy is not a power plant.
- Extreme, recurring capital needs — ARC will require billions more and serial dilutive rounds.
- Opaque financials (private) — no audited statements, no public price discovery.
- Key engineering risks (tritium fuel cycle, neutron/materials damage, disruptions) remain unproven at scale.
Opportunities
- Be first to a commercially relevant net-energy machine (Q>1) in 2027 — defines the category.
- Surging clean-firm / data-center power demand (AI) creating willing corporate offtakers and strategic investors.
- Independent HTS-magnet business line (fusion, MRI, grid, accelerators) as a near-term revenue hedge.
- US fusion policy support (DOE Milestone program, regulatory clarity under NRC byproduct-material framework).
- Technology licensing of the high-field-tokamak + magnet platform to other developers/nations.
Threats
- Competing fusion approaches reaching milestones first (Helion's pulsed FRC targeting 2028 power for Microsoft).
- Cheaper, sooner clean-firm alternatives — fission SMRs (Oklo, NuScale, X-energy), enhanced geothermal, solar+storage.
- A 'fusion winter' / capital-markets pullback freezing the next round before commercialization.
- Scientific shortfall: SPARC misses Q>1 or hits it far later/weaker than promised.
- Long-dated, binary outcome — a single materials/physics surprise can impair the whole thesis.
Moats, dependencies & bottlenecks
Moats
Proprietary design/manufacturing on Nobel-grounded high-temperature superconductors; demonstrated in hardware and DOE-validated. Shrinks reactor size/cost and is independently valuable beyond fusion.
~$6.85B raised — most of any private fusion company; outlasts under-funded rivals through a capital-intensive, long-dated build. Durability medium because the field needs serial multi-$B rounds.
Direct PSFC spinout with peer-reviewed ARC physics and a scientific talent pipeline; hard to replicate.
Google PPA, Eni $1B, DOE program, NVIDIA/Temasek/Mitsui — preferential access to demand, capital, and policy. Contingent on delivery.
Dependencies
Pre-revenue; entire program depends on continued access to large equity. A fusion-funding winter or rate shock could strand it.
Critical magnet input with a thin global supplier base; CFS is investing to scale it but remains exposed to capacity/cost.
The whole thesis gates on demonstrating net energy; physics/engineering surprises (disruptions, tritium, materials) could push it out or impair it.
Milestone-program funding and a workable byproduct-material regulatory path matter for timeline and cost.
PPAs are contingent on a plant that doesn't yet exist; cancellation/renegotiation risk if timelines slip.
Advantages
- Best-funded private fusion company globally (~$6.85B).
- Validated, proprietary 20-tesla HTS-magnet edge — the core enabling technology.
- Least-exotic, peer-reviewed high-field-tokamak approach with MIT physics pedigree.
- Tangible build progress (SPARC building + first TF magnet, Jan 2026) versus slideware rivals.
- Blue-chip demand + investor flywheel (Google PPA, Eni $1B, NVIDIA, Temasek, Mitsui, DOE).
Weaknesses
- Zero revenue; no commercial product; entirely equity-funded burn.
- Timeline already slipped (first plasma 2025→2027); scientific net energy ≠ a power plant.
- Extreme, recurring capital intensity — ARC needs billions more and serial dilution.
- Opaque private financials — no audited statements or public price discovery.
- Long-dated binary outcome — vulnerable to a single physics/materials surprise or a cheaper, sooner clean-firm competitor.
Bottlenecks
- SPARC must demonstrate net energy (Q>1) — the single gating scientific milestone; already slipped to 2027.
- HTS-magnet manufacturing must scale from demonstrator units to power-plant volumes reliably and affordably.
- Unsolved power-plant engineering — tritium breeding/fuel cycle, first-wall/divertor neutron damage, disruption mitigation at 20 T.
- Serial multi-$B financing — the ARC build is largely unfunded; capital availability gates commercialization.
- Regulatory/interconnection path for a first-of-kind fusion plant in Virginia.
Top signals & trends
Top signals
Sacra/press, May 2026. Capital depth is a structural advantage; exact post-money not disclosed.
CFS/Fortune, Jan 2026 — tangible build progress on the critical-path component.
Multiple sources, 2026 — schedule risk is the dominant near-term concern.
First-of-kind fusion offtake; demand-side validation, though contingent on plant delivery.
DOE/CFS, 2026 — federal validation and policy tailwind, modest dollar amount.
CFS/Wikipedia, 2026 — concrete site selection, but build is the next multi-$B unfunded step.
GeekWire/TechCrunch, 2026 — multiple well-funded rivals on different physics paths.
Inherent pre-revenue/private profile; drives LOW conviction.
Trends
Creates willing corporate offtakers (Google, Microsoft-Helion) and strategic investors (NVIDIA) for clean-firm baseload — the core demand thesis for fusion.
2025-26 saw CFS $3.85B, Helion $465M @ $15.5B, Pacific Fusion $1B+ — capital is flowing, favoring the best-funded incumbents.
DOE Milestone program and NRC byproduct-material framework reduce timeline/cost uncertainty for US developers.
Oklo, NuScale, X-energy, TerraPower can deploy clean-firm power before any fusion plant, competing for the same offtake.
First-plasma and net-energy dates slipping (CFS 2025→2027) keeps commercialization risk and 'fusion is always 10 years away' skepticism elevated.
Ecosystem & competitor graph
Suppliers feed the company; customers pull from it. Line thickness shows the strength of each tie (supply-chain dependency, customer earnings contribution). Hover to isolate a tie.
Fujikura, SuperPower/Furukawa, Shanghai Superconductor) High-temperature-superconducting tape is the critical input for the 20-tesla magnets; CFS has invested in scaling this supply chain. Non-US names listed as analysis only.
Investor (NVentures) and AI/simulation-compute supplier — referenced for AI-assisted plasma/engineering work alongside Siemens.
Industrial automation / digital-twin and engineering software partner for the SPARC build (analysis; ADR ticker).
Vacuum vessel, cryoplant, power electronics and precision machining suppliers for tokamak hardware — many private/not individually disclosed.
Offtaker — first-of-kind 200 MW corporate PPA on the ARC plant; also an equity investor (DeepMind partnership, Sept 2025).
Long-time investor and offtaker — committed ~$1B for power from CFS's first fusion reactor (Sept 2025).
Virginia utility / grid host — ARC sited in Dominion's Virginia service territory (James River Industrial Center); natural interconnection counterparty.
The structural demand pool (AI power crunch) CFS's PPAs and strategic investors target for clean-firm baseload.
Private. Pulsed FRC (non-tokamak) magneto-inertial approach; ~$15.5B valuation (Jun 2026, most valuable private fusion co.); Microsoft deal to supply power by 2028 — most aggressive timeline.
Private→public path. Field-reversed-configuration / p-B11 aneutronic fusion; ~$1.79B raised, announced Dec-2025 merger with Trump Media (TMTG) valuing combined co. ~$6B.
Private. Pulsed magnetic / inertial approach; launched with a $1B+ Series A — well-capitalized newer entrant.
Private (UK). Spherical-tokamak + HTS-magnet path — closest technological analog to CFS's high-field tokamak.
Public/intergovernmental megaproject (France). Conventional low-temperature-superconductor tokamak; slower and far costlier but the scientific reference program.
Public. Fast-fission microreactor — competes for the same clean-firm / data-center offtake demand on a sooner (late-2020s) timeline.
Public. NRC-certified light-water SMR — alternative clean-firm baseload that can deploy before any fusion plant.
Private (Gates-backed). Natrium sodium-cooled fast reactor under construction in Wyoming — advanced-fission alternative for firm clean power.