multimodal benchmark
Android Control Low_EM
Android control benchmark evaluating autonomous agents on mobile device interaction tasks with low exact match scoring criteria
Updated Aug 7, 2026
Published models3
Registry coverage3
MetricScore
EvidenceB
Android Control Low_EM leaderboard
Sorted by the source-provided rank. Higher score is better according to the registry.
| 1 | AC | 93.7% | 100.0% | 3 | C | |
| 2 | AC | 93.3% | 50.0% | 3 | C | |
| 3 | AC | 91.4% | 0.0% | 3 | C |
Score distribution
Top published rows on the benchmark's original scale.
Android Control Low_EM highlights
The top published results on this benchmark's own scale.
What is Android Control Low_EM?
Definition and scoring fields from the benchmark registry.
Android control benchmark evaluating autonomous agents on mobile device interaction tasks with low exact match scoring criteria
Scores are shown in ratio. The current registry marks this benchmark as not independently verified with evidence level B.
- Family
- Android Control Low_EM
- Modality
- multimodal
- Primary category
- multimodal
- Score direction
- higher
- LLMBoard eligible
- No
- Evaluation key
- android-control-low-em|llm-stats-current
Source-native results are preserved. Eligibility for the overall LLMBoard score is a separate policy decision.
FAQ
Common questions about Android Control Low_EM.
Which model scores highest on Android Control Low_EM?
Qwen2.5 VL 72B Instruct is currently ranked first with 93.7%.
What does Android Control Low_EM measure?
Android control benchmark evaluating autonomous agents on mobile device interaction tasks with low exact match scoring criteria
Is a higher score better?
Yes. Higher values rank better for this benchmark.
How many models are compared?
3 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.