Silvia Outperforms Every Major AI Model on Tax and Open Sources the Benchmark Test
Image: Primary ProCap Financial Inc. announced Monday that its AI agent Silvia outperformed every major AI model evaluated on tax-related topics and open-sourced the benchmark test used for the evaluation. The company said Silvia earned the highest overall score of 8.73 on a 10-point scale across ten expert-level tax scenarios covering federal law and five state tax codes, beating the next-best product, Claude Desktop, which scored 8.40.
The evaluation was conducted by a blind judge across four escalating levels of verification. Silvia retrieves answers from a proprietary library of primary tax law and was developed by a team of four AI engineers. Chairman and CEO Anthony Pompliano said the results address a theoretical debate among institutional investors about whether general-purpose model providers or domain-specific systems will capture more value in AI.
ProCap Financial is releasing the evaluation questions and framework so anyone can reproduce or contest the results. The company said Silvia now has more than $50 billion in assets connected to its platform across more than 20,000 users.
Sources
In this story
Published by Tech & Business, a media brand covering technology and business.
This story was sourced from AOL and reviewed by the T&B editorial agent team.
Back to Newswire
