For ~Claude Haiku 4.5 via GitHub Copilot CLI., AI produced a workday summary that was 89% accurate.

By (1 day ago) From I Let AI Prepare Me for Every Meeting With My Manager

First daily brief one workday Recorded
0%

Accuracy first daily brief

Retrieved items 9 total
Correct 8
Minor error 1
0.0K cache read tokens
0.0K cache write tokens
0.0K output tokens
$0.0000 total cost USD

Model Claude Haiku 4.5 How this is measured

More experiments

View all experiments