AI Code-Testing Startup Blacksmith Raises $45M at $550M Valuation

Blacksmith, an AI code-testing startup founded in 2024, raised $45 million in Series B funding led by Peak XV Partners, pushing the company’s valuation to $550 million. The round marks a stunning 9x increase from the $60 million valuation assigned just months earlier when the startup closed its Series A at $10 million. Existing investors GV and Y Combinator also participated in the new round, bringing total funding to $58.5 million.

The valuation surge reflects explosive momentum in a market where the pace of code generation has accelerated far beyond the pace of code validation. As AI coding assistants from OpenAI, Anthropic, and startups like Cursor have made writing software dramatically faster, the bottleneck has shifted downstream—to testing, verifying, and fixing that code before it reaches production.

“Validating code is still a bottleneck, and it’s an even bigger bottleneck because people are writing even more,” said Aditya Jayaprakash, Blacksmith’s co-founder and CEO, in an interview. The insight points to a structural imbalance in modern development: tools that accelerate creation have not been matched by tools that accelerate validation.

Nearly 10x Valuation Jump in Under a Year

Blacksmith’s trajectory from $60 million to $550 million in valuation within months reflects venture investor confidence in the broader thesis that code validation will become a standalone software category. The funding round sizes tell part of the story—a $10 million Series A followed quickly by a $45 million Series B—but the acceleration in valuation per dollar raised underscores how rapidly the market perceives this opportunity.

The startup’s timing aligns with a specific inflection point in software development. As companies deploy AI-assisted coding at scale, the volume of generated code grows, but so does the risk of unvalidated code reaching production. Blacksmith positioned itself to capture that demand shift, moving beyond a commodity infrastructure play into the emerging category of intelligent validation.

The Validation Bottleneck in AI-Accelerated Development

Read also: TechCrunch Disrupt 2026 targets the challenge of building in the AI era · June Raises $20M to Automate Enterprise AI Deployment · London’s anti-hustle hacker house challenges Silicon Valley founder culture · Fermi Appoints TerraPower Veteran as New CEO, Stock Surges 23%

Why Testing Became Critical

The rise of AI coding tools has created an asymmetry in the software development process. Cursor, OpenAI’s Codex, and Anthropic’s Claude Code can generate entire functions, modules, or features in seconds. Yet the quality of AI-generated code is inconsistent. Without equivalent acceleration in testing and validation, development teams face a choice: either validate every line manually, creating a new bottleneck, or ship unvalidated code at higher risk.

Blacksmith’s Series B announcement comes as this realization crystallizes across the industry. The startup’s largest customers include Mercury, Supabase, Clerk, Ashby, and Expensify—all infrastructure or developer-facing platforms where code quality is non-negotiable. These companies have adopted Blacksmith not as an experimental tool but as a core part of their release process, signaling that validation has graduated from a nice-to-have to a must-have.

The Gap Between Generation and Verification

The speed mismatch between generation and verification creates operational friction. A developer using Claude Code or Cursor can produce candidate code in seconds, but the testing and verification cycle—running continuous integration, checking code style, running security scans, and validating against automated test suites—still requires minutes to hours per build.

Blacksmith addressed this initially by offering cloud infrastructure optimized for continuous integration workloads, making the testing cycle itself faster. But the company has since added a new dimension: Codesmith, an AI coding agent that automatically fixes failed code checks. Rather than just running tests faster, Codesmith closes the loop by attempting to remediate failures autonomously, reducing iteration time from human-driven debugging cycles to automated repair cycles.

From 700 to 5,000 Customers in Months

The customer growth trajectory is perhaps the clearest signal of product-market fit. Blacksmith served more than 700 customers less than a year ago. Today it serves more than 5,000. That 7x expansion in customer count closely mirrors the 9x expansion in valuation, suggesting that growth is driven by real adoption rather than investor sentiment alone.

Revenue metrics underline the efficiency of that growth. Jayaprakash revealed that Blacksmith reached $10 million in annualized run rate with only 10 employees. The company has since grown its headcount to approximately 30 people and expanded revenue into what he described as “tens of millions of dollars.” He declined to provide an updated annual revenue figure, but disclosed that some of Blacksmith’s largest customers spend more than $1 million annually on the platform.

That customer concentration—where a handful of large accounts generate seven-figure annual contracts—is common in infrastructure software but typically requires significant enterprise sales effort. For a company that reached $10 million in annual revenue with 10 engineers, this pattern suggests the sales process is either highly efficient or driven by strong inbound demand as word-of-mouth spreads among engineering teams.

From CI Platform to AI-Powered Testing Agent

Blacksmith’s product roadmap reflects a strategic expansion beyond its initial narrow focus. The company started as a cloud provider for continuous integration workloads—essentially offering faster, cheaper infrastructure for running tests and builds. That remains the foundation, but Codesmith represents a departure into AI-powered remediation.

The distinction matters. A faster continuous integration platform is a commodity play competing on price and latency. An AI agent that fixes failures automatically is a higher-value tool because it reduces the human time required to debug and repair failed builds. This product evolution positions Blacksmith to expand into what Jayaprakash called “a broader suite of coding tools,” including not just testing but also writing, validation, and merging code—assembling a complete workflow platform rather than a point solution.

Competing Against GitHub, Cloud Giants, and Niche Startups

The Crowded Landscape

Blacksmith operates in a fragmented but increasingly crowded market. Direct competitors include GitHub Actions, which dominates continuous integration and continuous deployment by virtue of GitHub’s position as the default code repository platform. Cursor Automations brings code generation and validation under one umbrella. Both OpenAI’s Codex and Anthropic’s Claude Code include validation capabilities baked directly into their models.

Beyond startups, cloud infrastructure vendors—Amazon Web Services, Microsoft Azure, and Google Cloud—all offer code-testing and validation services, typically integrated with their broader cloud offerings. Each has existing customer relationships and can bundle testing validation as part of platform lock-in strategies, leveraging their dominant position in cloud infrastructure.

How Blacksmith Differentiates

Jayaprakash characterized Blacksmith’s strategy as competing on two dimensions: speed and affordability. Speed means faster test execution and faster failure remediation through AI-driven fixing. Affordability positions the platform as a cost-effective alternative to building custom continuous integration infrastructure or paying cloud vendor markups for validation services.

The data point that Jayaprakash highlighted—some customers spending more than $1 million annually—suggests that the addressable market is large enough to support multiple vendors. Customers spending seven figures on code validation are likely large organizations where the cost of failed validation, in shipping buggy code, exceeds the cost of premium tools. For such organizations, Blacksmith’s focus on acceleration and cost efficiency creates clear competitive separation from cloud vendor alternatives.

Building the Next Layer of the Coding Stack

Jayaprakash’s vision for Blacksmith extends beyond testing. The startup plans to expand into a complete coding workflow toolkit that spans writing, validating, and merging code—essentially becoming a platform layer that sits on top of repositories and above AI coding assistants.

This positioning makes strategic sense given the broader fragmentation in the AI coding category. Cursor handles generation and some validation. GitHub Copilot handles generation within integrated development environments. Claude Code handles generation for Claude users. But no single tool yet owns the full workflow from code generation to production deployment. Blacksmith is betting that a platform focused on the validation and remediation layer can become essential infrastructure for teams shipping AI-generated code at scale. The $550 million valuation and $45 million Series B funding represent investor bets that this thesis is correct—that validation will become a standalone, high-growth software category precisely because AI coding acceleration has made generation too fast for human validation alone.