Bloom Energy Corporation earnings recap
Q3 FY2026, reported October 22, 2026 · Beat 8 of last 8 quarters
Bloom Energy Corporation reported Q3 FY2026 revenue of $1.24B, a beat of 15.4% against consensus, and EPS of $0.94, a beat of 36.0%.
The Buildout is a website built by one person using AI agents, and could have factual errors.
Bloom's third guidance raise in a year, with product revenue above $1B in a single quarter and operating margin at 23.1%, is evidence that on-site fuel cell power is scaling as a genuine alternative to grid interconnection for AI data centers rather than a niche bridge. The Fremont expansion ahead of the 5 GW milestone, plus the AEP carbon-capture project and 800V DC work, suggests the constraint on deployment is shifting from customer readiness and permitting toward Bloom's own manufacturing capacity. Working capital consumed $27M of operating cash in the quarter as deliveries and inventory ramped — a pattern that may repeat as shipment volumes grow.
| Estimate | Actual | vs est | ||
|---|---|---|---|---|
| Revenue | $1.07B | $1.24B | +15.4% | beat |
| EPS | $0.69 | $0.94 | +36.0% | beat |
Bloom reported record third-quarter revenue of $1.24B, up 138% year-over-year, with product revenue crossing $1B for the first time at $1.07B and installation revenue of $84.9M up 66% sequentially. Operating income of $286M exceeded operating income for all of last year, and adjusted EBITDA was $299M (24% of revenue); non-GAAP diluted EPS was $0.94 versus $0.15 a year ago and GAAP EPS was $0.78. Product gross margin reached 38.4% even while ramping new production lines and expediting deliveries, while total gross margin of 34.5% was partly offset by installation mix. Cash flow from operations was a $27M use on working capital build in inventory, contract assets and receivables; capex was $153M in the quarter and $231M year to date, with free cash flow a $181M use in the quarter but positive $42M year to date, ending with $2.5B of cash. Commercially, the company cited a US Army conditional award to partner Indelible, AEP's 700 MW carbon-capture-enabled project for Google, first Texas shipments, and its first block of power ready at Nebius's New Jersey site two months after planning approvals.
Management raised full-year guidance for the third time in 2026, lifting revenue to $4.4–$4.6B (~122% growth at the midpoint), non-GAAP operating income to $960M–$1.03B (a 22% margin versus 14% at the midpoint of February's outlook) and EPS to $3.10–$3.40, while maintaining the roughly 34% gross margin outlook — up from 30% in 2025 and an original ~32% plan. The raise is built on two layers: backlog assessed customer by customer against site readiness and conversion timing, plus a probability-weighted pipeline of customers needing power within the quarter, with a small number of year-end deliveries determining where in the range they land. Capacity is being added incrementally rather than in step functions — the expanded Fremont footprint prepares the company for volume beyond its stated 5 GW annual output milestone, which management described as "just a milestone, in a long journey." KR framed CCS, 800V DC and combined heat and power as platform extensions riding the same fuel cell stack, called CCS a potential trillion-dollar opportunity, said hyperscalers are asking about carbon-capture readiness, and flagged expected international wins in the near future. Management emphasized a self-funded model, with operating income funding the 4x increase in capital spending while the cash balance stayed roughly in line with the start of the year.
“Our customer pipeline continues to grow and our backlog is continuing to expand faster than our revenue. With increasing diversity in customers, verticals, and geographies.”
on Backlog growth
“Operating margin for the quarter expanded to 23.1% from 8.9% a year ago. As revenue grew 138% while operating expenses grew only 26%.”
on Operating leverage
“Combustion technologies, do not offer a affordable and scalable pathway to carbon capture today. Bloom's technology in the commercial world is unique in that it can offer the concentrated carbon dioxide for carbon capture.”
on Carbon capture
Why expand the Fremont real estate footprint now, and should we assume capacity moves above 10 GW from the current 5 GW annual output once that facility is running?
KR said electricity demand is set to skyrocket across AI and many other segments globally, including for resiliency and security, and that the 5 GW milestone is clearly in sight, requiring preparation for what comes after. He stressed capacity will never be added in a step function but dialed up like an analog dial as need arrives — never past the company's headlights, but never so slowly that Bloom becomes the customer's bottleneck. No timeline was given for the expanded footprint.
Can you drill into the free cash flow profile — cash from operations was a $27M use despite margin expansion — and any puts and takes or one-offs to frame?
Simon Edwards said Bloom views itself as a self-funding business, with operating performance funding growth investment, reducing dependence on external capital and lowering execution risk. He noted the quarter's operating cash use is not a trend, since cash from operations is positive year to date even with higher people and capability investment and working capital funding for customer deployments, and pointed to full-year cash generation rather than any single quarter. He reiterated the expectation that the business will be free cash flow positive.
Where do the future initiatives — carbon capture, combined heat and power, and 800V DC — stand in terms of commercial opportunity and near-term market adoption?
KR called the AEP 700 MW carbon-capture project for Google a milestone, arguing combustion technologies lack an affordable, scalable path to carbon capture while Bloom's platform offers concentrated CO2, and sized the long-run opportunity at greater than a trillion dollars when the world is ready. On 800V DC he cited a company report showing a gigawatt data center saving over $5 billion in the first five years, said adoption is a question of "how soon" not "if," and noted conversations with all hyperscalers. He framed CHP as a natural add-on across AC or DC power, with or without carbon capture, particularly for commercial and industrial customers needing steam or cooling.