{
  "$schema": "https://themachinepress.com/schemas/story-v1.schema.json",
  "schema_version": "1.0.0",
  "document_type": "machine_press_story",
  "story": {
    "story_id": "mp-2026-09-26-014",
    "source_story_id": "tmp-story-jev-mobile-executor",
    "edition_id": "mp-2026-09-26-morning-0079",
    "edition_url": "https://themachinepress.com/edition/2026-09-26",
    "position": 14,
    "story_type": "dispatch",
    "section": "developer-tools",
    "editorial_classification": "editorial",
    "headline": "One Vision Plan Drove Several Phone Actions",
    "slug": "one-vision-plan-drove-several-phone-actions",
    "dek": "Jev-Mobile cut successful-run time by 32.7% and model API cost by 73.4% against a step-wise VLM baseline.",
    "summary": "Jev-Mobile cut successful-run time by 32.7% and model API cost by 73.4% against a step-wise VLM baseline.",
    "body_text": "Instead of asking a vision-language model to plan and ground every tap, Jev-Mobile uses infrequent VLM goals, the accessibility tree as an executable action space and a lightweight typed model for repeated local choices. On AndroidWorld it reached 79% task success, compared with 78% for SeeAct-V and 84% for the step-wise VLM baseline. The efficiency comparison includes only successful trajectories and depends on the tested mobile environment and serving prices.",
    "why_it_matters": "Jev-Mobile cut successful-run time by 32.7% and model API cost by 73.4% against a step-wise VLM baseline.",
    "limitations": [
      "The efficiency comparison includes only successful trajectories and depends on the tested mobile environment and serving prices."
    ],
    "importance": 8,
    "canonical_url": "https://themachinepress.com/story/mp-2026-09-26-014/one-vision-plan-drove-several-phone-actions",
    "json_url": "https://themachinepress.com/story/mp-2026-09-26-014.json",
    "first_published_at": "2026-09-26T09:00:00.000-04:00",
    "modified_at": "2026-09-26T09:00:00.000-04:00",
    "content_status": "new",
    "is_carryover": false,
    "carryover_reason": null,
    "key_claims": [
      {
        "claim_id": "claim-mp-2026-09-26-014-001",
        "text": "Jev-Mobile cut successful-run time by 32.7% and model API cost by 73.4% against a step-wise VLM baseline.",
        "source_ids": [
          "source-2026-09-26-014"
        ],
        "qualification": "The efficiency comparison includes only successful trajectories and depends on the tested mobile environment and serving prices."
      }
    ],
    "source_ids": [
      "source-2026-09-26-014"
    ],
    "tags": [
      "mobile agents",
      "GUI automation",
      "accessibility tree"
    ],
    "image_url": null,
    "corrections": []
  },
  "sources": [
    {
      "source_id": "source-2026-09-26-014",
      "title": "arXiv preprint 2609.30186",
      "publisher": "arXiv",
      "url": "https://arxiv.org/abs/2609.30186",
      "canonical_url": "https://arxiv.org/abs/2609.30186",
      "source_type": "primary_research",
      "is_primary_source": true,
      "published_at": "2026-09-24T13:30:32.000-04:00",
      "accessed_at": "2026-09-26T08:20:31.000-04:00",
      "supports_claim_ids": [
        "claim-mp-2026-09-26-014-001"
      ]
    }
  ],
  "corrections": [],
  "publisher": {
    "name": "The Machine Press",
    "url": "https://themachinepress.com",
    "description": "A daily newspaper for the age of artificial intelligence."
  },
  "cite_this_report": {
    "title": "One Vision Plan Drove Several Phone Actions",
    "publisher": "The Machine Press",
    "published_at": "2026-09-26T09:00:00.000-04:00",
    "canonical_url": "https://themachinepress.com/story/mp-2026-09-26-014/one-vision-plan-drove-several-phone-actions"
  }
}