Surge AI reports a 10.8% best pass rate on professional tax tasks
In its October 27 DAYJOB: Tax evaluation of 50 assignments, Claude Sonnet 5.5 leads when passing requires satisfying at least 95% of each task's criteria.
In its October 27 DAYJOB: Tax evaluation of 50 assignments, Claude Sonnet 5.5 leads when passing requires satisfying at least 95% of each task's criteria.