Agentic benchmark
AA Briefcase leaderboard
Artificial Analysis Briefcase. Every model the catalog carries a published AA Briefcase value for, ranked by that value.
CategoryAgentic
MeasureElo
TasksProfessional knowledge-work tasks
DifficultyProfessional work
An independently evaluated professional-work benchmark reported as Elo.
AA Briefcase ranking
15 models with a published AA Briefcase value, ordered by that value, highest first. Models the source has not scored on this benchmark are not listed — they are unmeasured, not last.
| Rank | Model | Provider | Elo |
|---|---|---|---|
| 1 |  Claude Opus 5 | Anthropic | 1720 |
| 2 | Claude Fable 5.1 | Anthropic | 1646 |
| 3 |  Muse Spark 1.3 | Meta | 1589 |
| 4 |  Qwen3.8-Flash-Next | Alibaba | 1587 |
| 5 |  GPT-6 Astra | OpenAI | 1562 |
| 6 |  Grok 4.6 | xAI | 1534 |
| 7 | Claude Fable 5 | Anthropic | 1530 |
| 8 |  GLM-5.3 | Z.AI | 1504 |
| 9 |  Kimi K3 | Moonshot AI | 1488 |
| 10 | GPT-5.6 Sol | OpenAI | 1463 |
| 11 | GLM-5.3-Flash | Z.AI | 1449 |
| 12 |  DeepSeek V4.1 Flash | DeepSeek | 1424 |
| 13 | Qwen3.8-27B | Alibaba | 1398 |
| 14 | GPT-5.6 Luna | OpenAI | 1333 |
| 15 | GPT-5.6 Terra | OpenAI | 1330 |
Evidence key: Observed
Rows are ordered by the value Artificial Analysis Briefcase Benchmark Leaderboard published, highest first. The source does not state whether a higher value is the better result, so this page does not either: for a benchmark that measures a rate of failure, read the table from the bottom.