AIReality checkStory 02
The clock cuts frontier agents down to 44%
What changedRealtime-Bench evaluates dynamic tasks where delayed actions change the outcome under actual wall-clock inference latency. The best wall-clock success was only 44 percent and best Pass^3 was 36 percent; extra reasoning had mixed effects under time pressure.
The useful part
Why it matters
The best wall-clock success was only 44 percent and best Pass^3 was 36 percent; extra reasoning had mixed effects under time pressure.
Keep in mind
Good to know
Coverage is limited to the benchmark's dynamic tasks.
Evidence