{
  "$schema": "https://themachinepress.com/schemas/story-v1.schema.json",
  "schema_version": "1.0.0",
  "document_type": "machine_press_story",
  "story": {
    "story_id": "mp-2026-10-18-019",
    "source_story_id": "tmp-brief-opengcua-stack",
    "edition_id": "mp-2026-10-18-morning-0100",
    "edition_url": "https://themachinepress.com/edition/2026-10-18",
    "position": 21,
    "story_type": "ticker",
    "section": "open-source",
    "editorial_classification": "editorial",
    "headline": "Computer Use Gets an Open Full-Stack Baseline",
    "slug": "computer-use-gets-an-open-full-stack-baseline",
    "dek": "OpenGCUA combines visual and programmatic interfaces, releases prompts and trajectories, and moves selected harness decisions into the model.",
    "summary": "OpenGCUA combines visual and programmatic interfaces, releases prompts and trajectories, and moves selected harness decisions into the model.",
    "body_text": "OpenGCUA defines a generalist computer-use agent that chooses how to observe, act and organize work across GUI and code interfaces. On OSWorld-v2, the full configuration improved mean partial score by 19.6 points and reduced reported steps by 53 percent across eight tested models relative to batched GUI-only execution. The team released implementation details, prompts and trajectories plus preliminary real-device and cross-device components. Four macOS tasks also exposed large model differences in reasoning-token use; the real-device evidence remains preliminary.",
    "why_it_matters": "OpenGCUA combines visual and programmatic interfaces, releases prompts and trajectories, and moves selected harness decisions into the model.",
    "limitations": [
      "On OSWorld-v2, the full configuration improved mean partial score by 19.6 points and reduced reported steps by 53 percent across eight tested models relative to batched GUI-only execution."
    ],
    "importance": 8,
    "canonical_url": "https://themachinepress.com/story/mp-2026-10-18-019/computer-use-gets-an-open-full-stack-baseline",
    "json_url": "https://themachinepress.com/story/mp-2026-10-18-019.json",
    "first_published_at": "2026-10-18T09:00:00.000-04:00",
    "modified_at": "2026-10-18T09:00:00.000-04:00",
    "content_status": "new",
    "is_carryover": false,
    "carryover_reason": null,
    "key_claims": [
      {
        "claim_id": "claim-mp-2026-10-18-019-001",
        "text": "OpenGCUA combines visual and programmatic interfaces, releases prompts and trajectories, and moves selected harness decisions into the model.",
        "source_ids": [
          "source-2026-10-18-021"
        ],
        "qualification": "On OSWorld-v2, the full configuration improved mean partial score by 19.6 points and reduced reported steps by 53 percent across eight tested models relative to batched GUI-only execution."
      }
    ],
    "source_ids": [
      "source-2026-10-18-021"
    ],
    "tags": [
      "computer use",
      "open source",
      "agents"
    ],
    "image_url": null,
    "corrections": []
  },
  "sources": [
    {
      "source_id": "source-2026-10-18-021",
      "title": "arXiv preprint 2610.22200, original v1",
      "publisher": "arXiv",
      "url": "https://arxiv.org/abs/2610.22200v1",
      "canonical_url": "https://arxiv.org/abs/2610.22200v1",
      "source_type": "primary_research",
      "is_primary_source": true,
      "published_at": "2026-10-15T12:51:35.000-04:00",
      "accessed_at": "2026-10-18T08:23:52.920-04:00",
      "supports_claim_ids": [
        "claim-mp-2026-10-18-019-001"
      ]
    }
  ],
  "corrections": [],
  "publisher": {
    "name": "The Machine Press",
    "url": "https://themachinepress.com",
    "description": "A daily newspaper for the age of artificial intelligence."
  },
  "cite_this_report": {
    "title": "Computer Use Gets an Open Full-Stack Baseline",
    "publisher": "The Machine Press",
    "published_at": "2026-10-18T09:00:00.000-04:00",
    "canonical_url": "https://themachinepress.com/story/mp-2026-10-18-019/computer-use-gets-an-open-full-stack-baseline"
  }
}