reasoning benchmark
AutomationBench
AutomationBench is a tool-use benchmark that evaluates AI agents on automating real-world workflows, testing their ability to orchestrate tools and complete multi-step automation tasks.
Updated Aug 7, 2026
Published models8
Registry coverage8
MetricScore
EvidenceB
AutomationBench leaderboard
Sorted by the source-provided rank. Higher score is better according to the registry.
| 1 | MA | 30.8% | 100.0% | 8 | C | |
| 2 | AN | 26.0% | 85.7% | 8 | C | |
| 3 | DE | 25.1% | 71.4% | 8 | C | |
| 4 | OP | 18.1% | 57.1% | 8 | C | |
| 5 | AN | 17.4% | 42.9% | 8 | C | |
| 6 | OP | 15.2% | 28.6% | 8 | C | |
| 7 | OP | 14.9% | 14.3% | 8 | C | |
| 8 | AN | 13.5% | 0.0% | 8 | C |
Score distribution
Top published rows on the benchmark's original scale.
AutomationBench highlights
The top published results on this benchmark's own scale.
What is AutomationBench?
Definition and scoring fields from the benchmark registry.
AutomationBench is a tool-use benchmark that evaluates AI agents on automating real-world workflows, testing their ability to orchestrate tools and complete multi-step automation tasks.
Scores are shown in ratio. The current registry marks this benchmark as not independently verified with evidence level B.
- Family
- AutomationBench
- Modality
- text
- Primary category
- reasoning
- Score direction
- higher
- LLMBoard eligible
- No
- Evaluation key
- automationbench|llm-stats-current
Source-native results are preserved. Eligibility for the overall LLMBoard score is a separate policy decision.
FAQ
Common questions about AutomationBench.
Which model scores highest on AutomationBench?
Kimi K3 is currently ranked first with 30.8%.
What does AutomationBench measure?
AutomationBench is a tool-use benchmark that evaluates AI agents on automating real-world workflows, testing their ability to orchestrate tools and complete multi-step automation tasks.
Is a higher score better?
Yes. Higher values rank better for this benchmark.
How many models are compared?
8 unique published model results are currently shown.
Does this benchmark affect the overall score?
This benchmark is preserved as source-native evidence but is not eligible for the current overall score.