AI 代码测试初创公司 Blacksmith 估值达 10 亿美元，AI 生成代码激增

Trendingnow1UnitedHealthcare and Humana are ending Medicare Advantage plans for nearly 1 million people2No wonder the White House wants Warsh to cut: A 1% Fed rate hike costs the Treasury an extra $95 billion a year in national debt interest3United Airlines CEO—who made $32 million last year—calls for a national wealth tax and a nearly 100% inheritance tax: 'I would pay them'1UnitedHealthcare and Humana are ending Medicare Advantage plans for nearly 1 million people2No wonder the White House wants Warsh to cut: A 1% Fed rate hike costs the Treasury an extra $95 billion a year in national debt interest3United Airlines CEO—who made $32 million last year—calls for a national wealth tax and a nearly 100% inheritance tax: 'I would pay them'AIFundingExclusive: AI code-testing startup Blacksmith hits $1 billion valuation as AI-written code surgesBy Beatrice NolanBeatrice NolanTech ReporterBy Beatrice NolanBeatrice NolanTech ReporterOctober 21, 2026, 7:00 AM ETBlacksmith founders. From left to right: Aayush Shah, Aditya Maru, Aditya (JP) Jayaprakash.Courtesy of Sequoia Capital/BlacksmithAdd Fortune on Google for similar content.

Blacksmith, a startup that runs the infrastructure companies use to test code before it ships, has been valued at $1 billion following a new $66 million funding round led by Sequoia Capital.Recommended Video

The round, which is Blacksmith’s Series C, comes about two months after the company raised a $45 million Series B led by Peak XV Partners at a $550 million valuation. It brings Blacksmith’s total funding to about $124.5 million.

Blacksmith says more than 10,000 companies now use its platform, up from about 5,000 in August. Customers include well-known startups Granola, Handshake, Supabase and LangChain. At the start of the year Blacksmith ran 5 million validation checks a week; it now runs 90 million. Annual recurring revenue has grown 17x over the same period, according to cofounder and CEO Aditya Jayaprakash, who said revenue is now in the tens of millions. The company said its largest customers spend more than $1 million a year.

Jayaprakash told Fortune that most of the new money will go toward purchasing more computing power. The company has about 40 employees across San Francisco and New York and plans to reach around 75 by the end of the year.

Blacksmith was founded in 2024 by Jayaprakash, and sells what developers call continuous integration, or CI: the automated builds and tests that code has to pass before it can be merged into an existing codebase and deployed. Most companies run those jobs on servers rented from AWS, Google Cloud or Microsoft Azure. Blacksmith instead buys its own machines, tens of thousands of them, according to Jayaprakash, and installs them in data centers around the world. It has writes its own software to run them, including systems that save the repetitive setup work each test run would otherwise redo from scratch.

A new bottleneck

Developers used to spend about half their time writing code and the other half waiting for it to build and pass checks, Jayaprakash said.

AI coding tools like Cursor, OpenAI’s Codex and Anthropic’s Claude Code have made the first half “extremely easy,” he said. “But now the bottleneck has shifted to the second piece, which is: how do I actually validate all this code?”

Even the leading labs are feeling the strain. Anthropic said last month that its own validation workloads grew 25 fold in six months as Claude took over most of its code output.

Blacksmith owns much of its own hardware—it buys its own servers rather than renting them from cloud giants like Amazon—which Jayaprakash says gives the company control over performance, but also exposes the company to the industry-wide shortage of compute and memory. Jayaprakash said capacity planning is one of the hardest parts of the business, because demand jumps every time a stronger model comes out. When Anthropic launched Claude Fable 5 in July, he said, CI volume on Blacksmith rose 40%.

As with many AI coding startups, there’s also a threat that leading AI model vendors—which already focus heavily on coding products—will eventually take over code validation themselves.

Coding agents like Codex and Claude Code can already run a project’s tests as they work, and Cursor launched a product in March that lets companies set up agents to automatically review and test new code. Meanwhile, Microsoft-owned GitHub sells its own hosted machines for running CI jobs. Jayaprakash argues Blacksmith is complementary to what the AI model companies do, because better models mean more code to check. “Usage on us actually goes up,” he said. Fortune Daily breaks the traditional barrier between audience and newsroom. The show transforms Fortune’s trusted reporting into actionable, conversational, and entertaining insights for an emerging class of business leaders. Watch here.About the AuthorBy Beatrice NolanTech Reporter

Beatrice Nolan is a tech reporter on Fortune’s AI team, covering artificial intelligence and emerging technologies and their impact on work, industry, and culture. She's based in Fortune's London office and holds a bachelor’s degree in English from the University of York. You can reach her securely via Signal at beatricenolan.08 See full bioAdd Fortune on Google for similar content.