Commonwealth Fusion Systems
MIT spinout building compact tokamak fusion reactors for grid power.
On the record
Checkpoint: SPARC first plasma and Q>1 net-energy-gain demonstration
The call
CFS leads private fusion on capital and construction progress, but the entire equity story waits on one unproven physics result due within a year.
Commonwealth Fusion Systems is building a demonstration nuclear fusion reactor called SPARC to prove that fusion can produce more energy than it takes to run. The company has raised billions of dollars and built most of the reactor, but nothing about it is proven yet, so there is no real competitor to compare it to or artificial intelligence system that could replace it. The entire company and all investor bets depend on whether SPARC actually achieves net energy gain when it first turns on in late 2026 or early 2027. If that test fails or gets delayed, the company might run out of time and money before proving the technology works.
Why
01
AI pressure: None apparent. AI data centers are a customer, not a predator: Google's 200MW PPA shows AI power demand feeding CFS, not eating it. Could AI do the core job? Unlikely.
02
Bigger players: Helion Energy. Helion's Microsoft power deal targets 2028 delivery; if it lands before CFS reaches the grid, CFS loses the 'first fusion power' narrative. Could a giant copy it? No, they are blocked.
03
What protects it. SPARC is ~75% built with magnets installed, plus signed PPAs with Google and Eni and an approved Chesterfield permit ahead of rivals.
What would change this call
SPARC achieves first plasma and reports a Q>1 net-energy-gain result (targeted late 2026/early 2027). Confirms or undermines the core physics claim the entire company and equity valuation depend on.
PJM Interconnection issues a decision on CFS's ARC 400MW interconnection application filed May 2026. Shows whether grid regulators are treating ARC as a real near-term power project or a speculative filing.
Helion Energy delivers or misses its 2028 power commitment to Microsoft. Would show whether a rival fusion startup beats CFS to an actual grid delivery milestone.
PJM Interconnection issues a decision on CFS's ARC 400MW interconnection application filed May 2026. Shows whether grid regulators are treating ARC as a real near-term power project or a speculative filing.
Helion Energy delivers or misses its 2028 power commitment to Microsoft. Would show whether a rival fusion startup beats CFS to an actual grid delivery milestone.
Key numbers
- $863M Series B2 round closed
- Google signs 200MW PPA for ARC plant
- First fusion PJM interconnection filing, 400MW ARC
Named risks
- SPARC net energy gain (Q>1) unproven, targeted late 2026/early 2027 with no fallback.
- Company's entire equity thesis depends on a single unproven physics milestone.
- US fusion plant licensing/regulatory framework still unsettled, could delay ARC.
- Reported $3.85B Series C figure unconfirmed by press release, funding picture uncertain.
- Rival fusion startups (Helion, Proxima) advancing quickly on funding and offtakes.
Rivals, as read in July 2026
| Rival | Direction | Evidence |
|---|---|---|
| Helion Energy | gaining on | Microsoft PPA for 50MW by 2028; raised $465M Series G at $15.5B valuation, June 2026 |
| Proxima Fusion | gaining on | Raised €411M in July 2026 led by XTX/East X, becoming Europe's best-funded fusion firm |
| TAE Technologies | contesting | Merging with Trump Media in ~$6B deal announced Dec 2025 to go public |
| General Fusion | contesting | Merging with Spring Valley to list on Nasdaq as GFUZ, first pure-play public fusion firm |
| China Fusion Energy Company | contesting | State-backed CNNC consolidated national fusion R&D into industrial vehicle, July 2025 |
Questions people ask
- Is Commonwealth Fusion Systems profitable?
- No. CFS is pre-revenue from electricity sales since its SPARC demo tokamak has not yet produced net fusion energy and its ARC commercial plant is unbuilt. The company funds operations through venture rounds, having raised more than $3 billion cumulatively as of its $863M Series B2 close on August 27, 2025.
- Has Commonwealth Fusion Systems had layoffs?
- No layoffs are reported. Headcount grew from about 643 employees in 2023 to roughly 1,137 by March 2026 per Revelio Labs, and CFS announced plans in January 2026 to expand its Devens, Massachusetts campus and add 100 or more jobs.
- Could AI replace Commonwealth Fusion Systems?
- No, its work is physical and regulatory, not software. CFS's moat is proprietary HTS magnet manufacturing, permits like its April 2026 PJM interconnection filing, and offtake contracts such as Google's 200MW PPA from July 2025. AI tools like Google DeepMind's TORAX instead help CFS accelerate its own plasma simulation work.
- Who owns Commonwealth Fusion Systems?
- CFS is privately held, spun out of MIT's Plasma Science and Fusion Center in 2018. Investors include Eni, Breakthrough Energy Ventures, Khosla Ventures, Temasek, Tiger Global (which led a $1.8B Series B in December 2021), NVIDIA's NVentures, and Google, which joined an $863M Series B2 in August 2025. There has been no IPO.
- Who is Commonwealth Fusion Systems' biggest competitor?
- Helion Energy is the closest rival, holding a Microsoft power purchase agreement targeting 50MW by 2028 and raising a $465M Series G in June 2026 at a $15.5B valuation. Proxima Fusion in Europe and TAE Technologies, which is merging to go public via a roughly $6B deal announced December 2025, are also contesting the space.
- When will Commonwealth Fusion Systems prove its fusion reactor works?
- CFS is targeting first plasma and a net energy gain (Q>1) demonstration on its SPARC tokamak in late 2026 or early 2027. As of mid-2026, SPARC construction was reported roughly 75% complete, with the first of 18 planned magnets installed.
Timeline
- 2026-07-16CFS named first international partner in UK UKAEA LIBRTI tritium blanket testing programme
- 2026-06-12CFS publishes five peer-reviewed papers detailing ARC reactor physics
- 2026-05-20A*STAR Singapore collaboration announced on ARC supply-chain technologies
- 2026-04-28CFS files first-ever fusion plant PJM interconnection application for 400MW ARC in Virginia
- 2026-03Headcount reaches roughly 1,137 employees per Revelio Labs
- 2026-01-12CFS announces plan to expand Devens campus, adding 100+ jobs
- 2025-09-19Chesterfield County, Virginia approves conditional-use permit for ARC plant
- 2025-09Eni signs power deal with CFS worth more than $1 billion
- 2025-08-27CFS closes $863 million Series B2 round, lifetime funding tops $3 billion
- 2025-07Google signs 200MW direct corporate power purchase agreement with CFS
Revision history
25 July 2026
First assessed. Losing its protection. The text above is fixed as of this date and is not edited afterwards.
In the same band
Sources
- Commonwealth Fusion Systems raises $863M Series B2 (2025-08-27)
- Commonwealth Fusion Systems funding, news & analysis
- Fusion Investment Hit a Record $4.48 Billion (2026-07-14)
- Commonwealth Fusion Systems installed the first of 18 magnets in its SPARC tokamak
- Wikipedia: Commonwealth Fusion Systems (2026-07-24)
- Going to the grid: CFS applies to plug ARC into PJM (2026-04-28)
- Chesterfield approves fusion power plant permit (2025-09-19)
- Revelio Labs: CFS employee count (2026-06-29)
- Crustdata: CFS headcount (2026-07-14)
- Commonwealth Fusion eyes major Devens expansion (2026-01-12)
- Six Companies Are Racing to Put Fusion on the Grid (2026-06-12)
- ZoomInfo: Commonwealth Fusion Systems Overview (2026-07-16)