Skip to main content
Global Edition•Technology
September 20, 2026•0 views
News Brief

AI Benchmarking Startup Vals Raises $40 Million Series A Led by Andreessen Horowitz

San Francisco-based AI benchmarking startup Vals raised $40 million in a Series A funding round led by Andreessen Horowitz.

AI Benchmarking Startup Vals Raises $40 Million Series A Led by Andreessen Horowitz
Image: TechcrunchCredit: Image: Techcrunch
Multi-Source Verification:Synthesized across 1 independent reporting sources.
single_source

In Brief

  • •San Francisco-based AI benchmarking startup Vals raised $40 million in a Series A funding round led by Andreessen Horowitz.
  • •Founded in 2024 by 25-year-old Stanford alumnus Rayan Krishnan, the startup evaluates AI models on complex, domain-specific tasks rather than abstract academic exams.
  • •Vals keeps its test materials undisclosed to prevent AI models from being trained directly on benchmark questions.
  • •The company's testing scope includes cybersecurity, law, finance, coding, biosecurity, and international humanitarian law.
  • •Vals reported an eightfold increase in revenue over the past year as businesses seek independent verification of AI performance.

Why It Matters

As tech companies deploy increasingly capable artificial intelligence, legacy academic benchmarks are struggling to measure real-world performance accurately. Furthermore, publicly available benchmark tests risk being compromised if AI developers train their models on test datasets. By offering private, task-based assessments across specialized sectors, third-party benchmarkers are becoming crucial gatekeepers in enterprise AI purchasing decisions.

Key Details

Series A Funding
$40 million
Lead Investor
Andreessen Horowitz
Seed Investors
8VC and Bloomberg Beta
Founding Year
2024
Revenue Growth
8x year-over-year

Context & Background

Historically, artificial intelligence capability was evaluated using open academic exams, such as standardized tests or bar examinations. As frontier models advance rapidly, industry observers have noted that models can memorise or train on these public materials, necessitating private evaluation systems that assess real-world application quality.

Original Reporting & Sources

VNews is an independent briefing product. Original reporting remains credited to the following publishers:

Techcrunch
Primary Source
Read Original ↗
About VNews AI News Briefs

This brief was synthesized automatically using attributed third-party reporting. VNews is a digital news publication produced under the VContent Group brand and operated by Francis Vincent Sunday in Ghana, independent from the publishers cited above. Read our AI Transparency Policy and Editorial & Corrections Policy.

Related Stories

Explore More