BREAKING
AA-Briefcase Benchmark Launches
Multi-week knowledge-work tasks
1
4 scenarios
↓
2
91 tasks
↓
3
Real deliverables
AA-Briefcase Elo leaderboard
Claude Fable 5
1587
Opus 4.8
1356
GLM-5.2
1266
GPT-5.5
1159
0
B
total params
0
B
active params
0
M
context tokens
0
x
vs GLM-5.1
0
x
vs Kimi K2.6
0
%
tasks fully solved
Real knowledge-work stays hard
AI NEWS BLITZ
Artificial Analysis just launched AA-Briefcase, a new test for real knowledge-work.