Benchmarks › AA-Briefcase
1,817top score · Claude Sonnet 5.5
140models ranked
83open-weights models
Crowd-BT Elohigher is better
Oct 20, 2026last updated
How the test works
1
Prompts
Produce professional files for multi-step workplace scenarios
- Kinds of tasksProfessional work
- In → outText, files → Files
2
Votes
Pairwise comparisons fitted with Crowd-BT (v1.1 Briefcase / v2.1 GDPval)
3
Rating
Crowd-BT Elo
- Better isHigher
Sources: artificialanalysis.ai
AA-Briefcase 1.1 leaderboardtop 10 of 140
Full leaderboard with filters →| Model | AA-Briefcase 1.1 | Per task | Price | Speed | Supports | Context |
|---|---|---|---|---|---|---|
1,817max | $5.46 | $2 / $10 | 137 | 1M | ||
1,803max | $5.98 | $4 / $20 | 96 | 1M | ||
1,729 | — | — | 1M | |||
1,672max | $7.63 | $10 / $50 | 69 | 1M | ||
1,659max | $5.86 | $5 / $25 | 55 | 1M | ||
1,659max | $4.52 | $4 / $20 | 161 | 1M | ||
1,642xhigh | $3.74 | $2 / $6 | 94 | 500K | ||
1,613thinking | $5.41 | $2 / $6 | 40 | 984K | ||
1,581max | $1.60 | $1.25 / $4.25 | 128 | 1M | ||
1,580thinking | $0.37 | $0.15 / $0.47 | 55 | 262K |
AA-Briefcase questions
What is AA-Briefcase?›
Produce professional files for multi-step workplace scenarios. It has 91 tasks. Created by Artificial Analysis.
Which model is best on AA-Briefcase?›
Claude Sonnet 5.5 leads AA-Briefcase 1.1 with 1,817, ahead of Claude Opus 5.5 (1,803) and Muse Wonder Full (1,729).
What is the best open-weights model on AA-Briefcase?›
Qwen3.8 Flash with 1,580, then DeepSeek V4.1 Flash 1017 (1,570).
How is AA-Briefcase scored?›
Crowd-BT Elo; pairwise comparisons fitted with crowd-bt (v1.1 briefcase / v2.1 gdpval); 1 run per task per task.