Blacksmith Hits $1B as Weekly CI Jobs Jump From 5M to 90M

A Fortune report published October 21, 2026, says Blacksmith raised a $66 million Series C led by Sequoia Capital at a $1 billion valuation; co-founder and CEO Aditya Jayaprakash framed the pressure behind the investment as “how do I actually validate all this code?” The company expects to spend most of the new money on computing power for the automated builds and tests it runs for software teams.

Blacksmith’s Series C announcement says weekly continuous-integration jobs rose from 5 million in January to more than 90 million, annual recurring revenue grew 17-fold, and more than 10,000 companies use the platform. It names Premji Invest and existing investors Peak XV Partners, GV and Y Combinator as participants alongside Sequoia, and puts total funding at a rounded $125 million. Those workload, revenue-growth and customer figures are company-reported operating claims.

The August valuation followed a March financing

Blacksmith’s August 12 Series B disclosure says the company raised $45 million at a $550 million valuation in March 2026, with Peak XV Partners leading the round. The earlier financing was made public in August, leaving a much shorter gap between announcements than between the rounds themselves. On the stated valuations, the October price is about 82% higher than the March price.

That increase reflects what investors were willing to pay for a stake as Blacksmith expanded its workload and prepared to add capacity. It does not measure profit or cash generation. The $125 million raised to date is also a cumulative financing figure, rather than the amount available for the next buildout. Blacksmith has disclosed a growth multiple for annual recurring revenue, but not the starting and current dollar amounts needed to compare that revenue with the valuation.

The financing sequence matters because the business must acquire capacity before it can sell all the work that capacity might handle. A higher valuation can support another large round, while the operating question remains whether paid demand will occupy enough of the added machines. Job volume establishes the scale of activity on the platform; it does not answer that financial question on its own.

AI coding raises the volume of work that CI must absorb

Continuous integration runs automated builds and tests on code changes before those changes are merged. A coding agent can produce changes quickly, but each proposed change still needs validation. Blacksmith’s pitch to investors is that faster code creation increases demand for the computing infrastructure that performs those checks. The reported rise in jobs is consistent with heavier demand on its platform, though the total does not show how many jobs were directly caused by AI-written code.

Weekly jobs are a useful measure of throughput, but they are not interchangeable with computing demand. One job might finish quickly; another might run a larger test suite or require a more powerful machine. A higher count can therefore increase the burden on the fleet without increasing it in exact proportion. It can also change the mix of machines Blacksmith needs to have ready when customers submit work.

Timing adds another constraint. CI demand arrives when developers push changes, and customers may need many jobs to run at once before work can move forward. Blacksmith describes individual customers’ demand as swinging sharply as batches of builds and tests begin and end. To keep queues short during a surge, the company needs machines available before those jobs arrive; during quieter periods, some of that capacity may sit idle.

The reported 17-fold rise in annual recurring revenue points to commercial growth alongside the workload increase, but the two measures answer different questions. Job counts describe activity, while ARR describes the annualized value of recurring contracts. Without disclosed revenue totals, job durations or utilization, the public figures cannot show how efficiently Blacksmith converts the additional work into revenue.

More owned compute makes growth expensive upfront

Blacksmith buys servers and installs them in data centers instead of relying solely on rented general-purpose cloud machines for its CI workloads. Owning equipment gives it direct control over how that capacity is configured and operated, but requires spending on hardware ahead of future customer demand. The company’s plan to expand its fleet gives the financing a concrete use: it needs machines, power and data-center space to carry more build-and-test work.

Its current footprint is just under 10 MW across data centers in the United States and Europe, with a target to grow it at least fivefold in 2027. That target describes planned capacity, not machines already installed. Adding servers also takes more than a purchase order: the company has to secure equipment, place it where power and space are available, and bring it into service as demand grows.

That creates a timing risk on both sides. Too little available capacity can leave customers waiting for tests and builds; too much capacity bought early can leave expensive machines underused. Jayaprakash has also identified shortages of compute and memory as a complication for a company purchasing its own hardware. The round gives Blacksmith capital to pursue the expansion, while the return on that spending will depend on sustained paid use of the larger fleet.

The next concrete milestone is the planned 2027 capacity expansion. Its significance will depend on how much of the added compute carries customer workloads and what those workloads earn after hardware and data-center costs. More precise revenue, utilization and cost disclosures would make that relationship visible; for now, the company’s job and ARR growth figures show demand without establishing the economics of the planned fleet.

Also read:

• Nous Raises $90M—A $1.5B Valuation Meets a $36M Revenue Run Rate
• Netflix Cuts Nearly 1,000 Jobs While Profits Are Still Growing
• Suki Raises $90M—Its Next AI Agents Move Beyond Clinical Notes