{
  "id": "benchmark:android-world",
  "_kind": "Benchmark",
  "_file": "benchmarks/benchmarks/benchmarks-agentic-fullstack.yaml",
  "_cluster": "benchmarks",
  "attributes": {
    "displayName": "AndroidWorld",
    "homepageUrl": "https://google-research.github.io/android_world/",
    "kind": "full-stack",
    "targetsKind": "AgentVersion",
    "description": "AndroidWorld (Google Research, 2024) is a dynamic Android\nenvironment benchmark of 116 tasks across 20 real apps, used to\nevaluate autonomous mobile-UI agents on natural-language goals\nwith stochastic real-app state.\n"
  },
  "outgoingEdges": [
    {
      "from": "benchmark:android-world",
      "to": "skill-area:android-native",
      "kind": "covers",
      "attributes": {
        "attributes": {
          "coverage": "tangential",
          "weight": 0.2
        }
      }
    }
  ],
  "incomingEdges": [
    {
      "from": "eval-result:android-world.gemini-2-5-pro.001",
      "to": "benchmark:android-world",
      "kind": "scored_against",
      "attributes": {}
    },
    {
      "from": "eval-run:android-world.gemini-2-5-pro.2025-06",
      "to": "benchmark:android-world",
      "kind": "for_benchmark"
    }
  ]
}