Blacksmith has raised $45 million in Series B funding as the growing use of AI coding tools increases the volume of software changes that development teams need to build, test and validate.
The San Francisco-based infrastructure company announced the funding on August 12, 2026. The round was led by Peak XV Partners, with existing investors GV and Y Combinator also participating. Blacksmith said the financing gives the company a valuation of $550 million and brings its total funding to $58.5 million.
The company provides infrastructure for continuous integration, the part of the software development process where code changes are built and tested before they are merged. Blacksmith says the rapid adoption of coding agents is increasing the number of pull requests that development teams generate, creating greater demand for the computing resources used to validate those changes.
Blacksmith Chief Executive Officer Aditya Jayaprakash said customers using coding agents are generating several times more pull requests, increasing the load placed on systems responsible for checking whether those changes work as intended.
Blacksmith said CI jobs running on its infrastructure have increased by between 5% and 10% week over week since the beginning of 2026. The company currently operates hundreds of thousands of computing cores and plans to increase its computing capacity by roughly 10 times in the coming months.
Most of the new funding will go toward expanding that computing capacity, according to the company. Blacksmith is also hiring engineers in New York and San Francisco as it expands its infrastructure and products.
The increase in software testing demand is linked by Blacksmith to the growth of AI-assisted development. Tools such as Anthropic’s Claude Code and OpenAI’s Codex can generate and modify software more quickly than conventional manual development, but the resulting code still has to pass builds and tests before it can be incorporated into a project.
Blacksmith’s business is built around that validation process. The company provides infrastructure designed to run GitHub Actions workloads, using dedicated computing resources, caching and storage optimized for software builds and tests.
The company’s platform uses bare-metal machines and micro-virtual-machine isolation for CI workloads. Blacksmith has also developed tooling for analyzing GitHub Actions performance, including information on pipeline speed, failures, flaky jobs and CI spending.
Blacksmith’s technical work extends to software logging as well. The company has described using ClickHouse to store and analyze GitHub Actions logs, allowing development teams to search large volumes of build and test data to investigate failures and other problems.
The company is also expanding beyond infrastructure with an AI coding product called Codesmith. Blacksmith describes Codesmith as a cloud coding agent designed to work within the software development and testing process.
According to Blacksmith, Codesmith can diagnose CI failures and make fixes when checks fail. The company is also developing Codesmith QA to test changes autonomously before they are merged.
That development reflects a broader direction in Blacksmith’s products: using software agents not only to produce code, but also to help maintain the systems that determine whether the code is ready to be accepted.
Blacksmith has been expanding quickly since its earlier funding rounds. In May 2025, the company announced a $3.5 million funding round led by GV and Y Combinator. At the time, it said its infrastructure was being used by more than 600 organizations.
Blacksmith followed that round with a $10 million Series A in September 2025, again led by GV. Contemporary reporting from TechCrunch said the company had reached $3.5 million in annual recurring revenue and more than 700 customers at the time.
Blacksmith now says it serves more than 6,000 companies. It lists Supabase, Clerk, Ashby and Mercury among its customers.
The company has also expanded its infrastructure beyond conventional Linux workloads. In April 2026, Blacksmith announced the general availability of M4-powered Mac runners, aimed at teams that need Apple hardware for their build and testing processes.
Blacksmith said some customers testing the M4 infrastructure had seen builds run more than twice as fast, attributing the improvement to the newer hardware and its infrastructure design. Those performance figures are company-reported results rather than independent benchmarks.
The company’s earlier work also included using coding agents internally to accelerate the development of its Mac runners. Blacksmith described a development process in which an agent could make a change, run tests, examine the result and continue iterating without requiring a developer to intervene at every stage.
Blacksmith’s latest financing therefore comes as the company expands both the physical infrastructure used to run software tests and the AI systems intended to help diagnose and fix failures within those tests.
The company argues that as AI systems generate more software changes, the infrastructure responsible for checking those changes must scale with them. Its planned 10-fold increase in computing capacity is aimed at handling the additional CI workload it expects as coding agents become more widely used.
Source: Blacksmith Series B funding announcement.
Discover more from Aree Blog
Subscribe now to keep reading and get access to the full archive.
