Skip to main content

AndroidWorld

A mobile GUI agent benchmark for completing Android app workflows and on-device tasks.

Benchmark score on AndroidWorld — June 2, 2026

BenchLM mirrors the published score view for AndroidWorld. Holo3.1-35B-A3B leads the public snapshot at 79.3% , followed by Holo3.1-4B (71.0%) and Holo3.1-9B (71.0%). BenchLM does not use these results to rank models overall.

4 modelsAgenticCurrentDisplay onlyUpdated June 2, 2026

The published AndroidWorld snapshot is tightly clustered at the top: Holo3.1-35B-A3B sits at 79.3%, while the third row is only 8.3 points behind. The broader top-10 spread is 9.0 points, so many of the published scores sit in a relatively narrow band.

4 models have been evaluated on AndroidWorld. The benchmark falls in the Agentic category. This category carries a 22% weight in BenchLM.ai's overall scoring system. AndroidWorld is currently displayed for reference but excluded from the scoring formula, so it does not directly affect overall rankings.

About AndroidWorld

Year

2026

Tasks

Android app workflows

Format

Interactive mobile-agent evaluation

Difficulty

Complex mobile task completion

BenchLM tracks AndroidWorld as a display-only benchmark reference when providers publish exact values alongside broader GUI-agent summaries.

BenchLM freshness & provenance

Version

AndroidWorld 2026

Refresh cadence

Quarterly

Staleness state

Current

Question availability

Public benchmark set

CurrentDisplay only

BenchLM uses freshness metadata to decide whether a benchmark should still be treated as a strong differentiator, a benchmark to watch, or a display-only reference. For the full scoring policy, see the BenchLM methodology page.

Benchmark score table (4 models)

1
79.3%
2
71.0%
3
71.0%
4
70.3%

FAQ

What does AndroidWorld measure?

A mobile GUI agent benchmark for completing Android app workflows and on-device tasks.

Which model scores highest on AndroidWorld?

Holo3.1-35B-A3B by H Company currently leads with a score of 79.3% on AndroidWorld.

How many models are evaluated on AndroidWorld?

4 AI models have been evaluated on AndroidWorld on BenchLM.

Last updated: June 2, 2026 · BenchLM version AndroidWorld 2026

The AI models change fast. We track them for you.

For engineers, researchers, and the plain curious — a weekly brief on new models, ranking shifts, and pricing changes.

Free. No spam. Unsubscribe anytime.