AutomationBench-AA
Explore score history ↗AutomationBench-AA is Artificial Analysis's independently run version of AutomationBench, covering 657 real-world SaaS workflow tasks across 40 simulated applications (e.g. Gmail, Slack, Salesforce, HubSpot). It scores the share of objectives an agent completes without violating business guardrails, using a private held-out task set.
- Modality
- text
- Evaluates
- reasoning, agents, tool calling
- Openness
- unknown
- Importer
- LLM Stats
Source provenance
AutomationBench-AA results and reported scores
Reported benchmark scores · LLM Stats
Reported scores: 1
| Model | Organization | Reported value | Reported | Evidence |
|---|---|---|---|---|
| Grok 4.5 | xAI | 0.514 | 2026-07-16 | evidence |
Scores are partitioned by the source that reported them and are never merged into a single cross-source ranking, because the sources measure different things and say so.