Skip to content

openai, anthropic, google, meta and xai trust one startup to grade their models

pulse Illustration of figure stamping documents at a desk with OpenAI, Anthropic, Google, Meta, and xAI logos below

top ai rounds this week – tracked by our ai host nathan on capital radar:

$143m – @coderabbitai series c at $1.5b. reviews code that ai agents write before it ships to production. 2m+ reviews per week, 17,000 customers, revenue up 5x yoy. atomico and smash capital co-led.

$60m – @CormaAI emerged from stealth with a seed led by sequoia, khosla and coatue. a foundation model that only does one thing – defend. in corma's own simulations, ai attackers won 88% of the time. defenders caught 12%.

$40m – @ValsAI series a at $400m, a16z leading. independent benchmarks that test frontier models on real tasks in law, finance and coding. every major lab cites the scores in official model cards. revenue up 8x since 2025.

$30m – @mindgard series a, album vc. attacks ai systems to find vulnerabilities before real attackers do. 150+ disclosed flaws, including a zero-day in cursor and bypassed guardrails in chatgpt.

$20m – @attestablecom closed a seed co-led by altimeter and tlv partners. cryptographic proof that a model ran the right computation on the right data – without exposing the model itself. cofounders from starkware and safe superintelligence.

the pattern: code review, model eval, red-teaming, cryptographic verification, defensive cybersecurity – five different answers to the same question: how do you trust ai at production scale?

Weekly Capital Radar table showing five AI funding rounds: CodeRabbit $143m, Corma $60m, Vals AI $40m, Mindgard $30m, Attestable $20m

Stay in the loop

Get the latest AI news delivered to your inbox weekly

Thanks for subscribing!