Startup Vals Secured $40 Million in Funding
The San Francisco-based AI evaluation firm raised capital to expand its team and operations.
Updated on Sept. 19, 2026 in Startups
Live Poll
Should AI companies be required to use independent benchmarking to prove their performance claims?
In August 2026, the San Francisco startup Vals successfully raised $40 million in a series A funding round led by Andreessen Horowitz. The company provides model evaluation services to federal agencies and private firms to measure the performance and capabilities of AI systems.
Why it matters
Existing academic benchmarks struggle to keep pace with rapid advancements in AI, creating a need for specialized firms to verify model reliability. Vals addresses this by testing AI performance across complex domains like law, finance, and coding to identify potential weaknesses.
The firm reports revenue that is eight times higher than last year and has expanded its staff from eight to 25 employees. Vals currently plans to add 10 to 15 additional staff members as it prepares to relocate to a larger office.
The players
Vals
This San Francisco-based startup provides specialized evaluation services for artificial intelligence models.
Andreessen Horowitz
This prominent Silicon Valley venture capital firm led the latest series A funding round for the startup.
8VC
This venture capital firm was a lead investor in the startup's previous seed funding round.
Bloomberg Beta
This venture capital organization contributed to the startup's initial seed investment stage.
The details
Companies engage Vals to rigorously test their models for task performance, ensuring verification of capabilities and potential negative impacts. The startup is currently headquartered on Folsom Street in San Francisco.
Timeline
Vals was formed in 2024.
The company secured a seed funding round in 2025.
The series A funding of $40 million was raised in August 2026.
A reporter visited the company office in September 2026.
Market Landscape
This investment highlights the increasing capital flowing toward AI infrastructure and quality assurance firms. Vals positions itself against traditional academic benchmarking by offering proprietary performance testing to commercial and federal entities.
While the funding does not change retail prices, it signals that organizations will have more robust data regarding the reliability of the AI tools they integrate into their products. Clients in law, finance, and coding can expect more standardized performance metrics as the firm scales its operations.
The takeaway
As AI models grow in complexity, the ability to independently verify their performance is becoming a critical business service. Investors are increasingly backing firms that provide this layer of professional auditing to ensure AI systems function safely across high-stakes industries.
Further reading
For more on the local innovation ecosystem, explore the latest news in San Francisco Startups.
Live Poll
Should AI companies be required to use independent benchmarking to prove their performance claims?
