Blacksmith raised $45 million in Series B funding on 12 August 2026, led by Peak XV Partners with participation from GV and Y Combinator, pushing its valuation to $550 million — a roughly tenfold jump from the $60 million it was worth at its September 2025 Series A. The company’s pitch is narrow and specific: build, test and verify software before it reaches production, positioned squarely as the validation layer sitting underneath the growing volume of code written by Claude Code, Codex and other AI coding agents. Customer count has grown from around 800 at Series A to more than 6,000 today, including Supabase, Clerk, Ashby and Mercury, with CI job volume on its infrastructure climbing 5–10% week over week since the start of 2026.
The number worth sitting with is the growth rate of the underlying problem, not just the funding round. As AI coding agents let teams generate pull requests faster than any human review process was designed to handle, the bottleneck in software delivery is quietly shifting from “can we write this code” to “can we trust this code enough to ship it.” Investors backing a pure-play validation infrastructure company at this scale is a signal that the market now sees code review and CI capacity — not code generation — as the constraint that actually limits how fast AI-assisted teams can move.
So what
If your team has adopted AI coding tools and shipping velocity has gone up, it’s worth asking honestly whether your review and testing capacity scaled with it, or whether pull request volume just quietly outran what your process can safely validate. Faster generation without a matching investment in validation is exactly the gap that turns AI-assisted prototypes into production incidents. We build AI-assisted development workflows with proper review, testing and provenance built in from the start, not bolted on afterwards — see our AI-assisted development approach or get in touch to talk through where your process needs reinforcing.