Code Metal raises $125M at $1.25B valuation for defense AI

Code Metal raises $125 million at $1.25 billion valuation

Boston-based startup Code Metal has raised a $125 million Series B funding round, valuing the company at $1.25 billion. The round was led by Salesforce Ventures and included Accel, B Capital, and J2 Ventures.

This comes just months after the company closed a $36 million Series A led by Accel. Code Metal, founded in 2023, uses AI to translate and verify code, with a focus on the defense industry.

The AI code translation gold rush

Code Metal is part of a wave of startups selling tools for the AI coding boom. Over the past two years, companies like Antithesis, Code Rabbit, and Harness have also secured millions to automate, test, and secure AI-generated code.

These companies are the "picks and shovels" for an industry increasingly reliant on AI. While their methodologies are often unproven, investors are betting that some will succeed as the market matures.

Code Metal's core problem is translating old, legacy code into modern languages. CEO Peter Morales, a former Microsoft and MIT Lincoln Laboratory employee, says this is a major bottleneck, especially for government and defense contractors with code written in obsolete languages.

Selling to defense and navigating bugs

The startup has secured early customers in sensitive sectors. Its client list includes L3Harris, RTX, and the US Air Force. It is also working with Toshiba and is in talks with a major, unnamed chip company.

Its platform translates code from high-level languages like Python and C++ to lower-level or hardware-specific languages like Rust, VHDL, and Nvidia's CUDA. The big risk in this process is introducing catastrophic bugs.

"In the course of translation, you might be inserting bugs—which is catastrophically problematic," said Yan-David Erlich, a general partner at investor B Capital.

Code Metal claims its proprietary tech mitigates this by generating test harnesses at each translation step to verify the code works. Morales states that for its current pipelines, the software cannot generate an erroneous translation; it will simply state no solution is possible if it fails.

A new model for pricing AI software

The company is navigating how to price its AI enterprise software. It is moving away from the traditional "per seat" model used by giants like Oracle and Salesforce.

Instead, Code Metal negotiates pricing individually with each customer based on one of three metrics:

• The time to develop a kernel.
• Lines of code translated (based on time to write code).
• Development time saved.

Morales admits this value-based pricing can "get murky," but claims every pilot deployment has moved to a next phase. The model charges for the time saved, not just the raw cost of the AI tokens processed.

Building a sales team for enterprise growth

To support its sales strategy, Code Metal has hired key executives from the enterprise software world. Ryan Aytay, former CEO of Tableau (owned by Salesforce), is now its President and COO.

Last year, the company also brought on Laura Shen, previously the director for China at the US National Security Council, as its executive vice president of growth.

With its new funding and $1.25 billion valuation, Code Metal claims it is already profitable. This makes it a rare "unicorn" that is both highly valued and generating positive cash flow from the start.

Whether its focus on defense and complex code translation translates to long-term success in the crowded AI landscape remains the ultimate test.