Claims face scrutiny

Customers and independent researchers probe whether the 99% GPT-6 accuracy claim holds across domains beyond TypeSafe’s preferred benchmark. Some buyers pause migrations until Jev’s eval reliability is proven in regulated and safety-critical workflows.

Each figure is the modeled move of the S&P 500 if the outcome occurs. For information only. Not investment advice.