{
  "$schema": "https://themachinepress.com/schemas/story-v1.schema.json",
  "schema_version": "1.0.0",
  "document_type": "machine_press_story",
  "story": {
    "story_id": "mp-2026-10-31-008",
    "source_story_id": "tmp-story-context-energy-cost",
    "edition_id": "mp-2026-10-31-morning-0113",
    "edition_url": "https://themachinepress.com/edition/2026-10-31",
    "position": 8,
    "story_type": "dispatch",
    "section": "infrastructure",
    "editorial_classification": "editorial",
    "headline": "The Input Context Has an Energy Bill Too",
    "slug": "the-input-context-has-an-energy-bill-too",
    "dek": "Profiling found linear output scaling but both linear and quadratic components for long-input prefill compute time and energy.",
    "summary": "Profiling found linear output scaling but both linear and quadratic components for long-input prefill compute time and energy.",
    "body_text": "Energy estimates for language models often count generated tokens and neglect the long prompts common in agent workflows. A profiling study varied both prefill and decode lengths and fit generalized linear models to compute time and measured energy. Decode cost scaled linearly with output length, while prefill showed linear and quadratic components as input grew; compute time also tracked energy closely enough to serve as a proxy in the tested setup. The work offers an estimation framework, not a measurement of every proprietary API or deployment stack.",
    "why_it_matters": "Profiling found linear output scaling but both linear and quadratic components for long-input prefill compute time and energy.",
    "limitations": [
      "The work offers an estimation framework, not a measurement of every proprietary API or deployment stack."
    ],
    "importance": 9,
    "canonical_url": "https://themachinepress.com/story/mp-2026-10-31-008/the-input-context-has-an-energy-bill-too",
    "json_url": "https://themachinepress.com/story/mp-2026-10-31-008.json",
    "first_published_at": "2026-10-31T09:00:00.000-04:00",
    "modified_at": "2026-10-31T09:00:00.000-04:00",
    "content_status": "new",
    "is_carryover": false,
    "carryover_reason": null,
    "key_claims": [
      {
        "claim_id": "claim-mp-2026-10-31-008-001",
        "text": "Profiling found linear output scaling but both linear and quadratic components for long-input prefill compute time and energy.",
        "source_ids": [
          "source-2026-10-31-008"
        ],
        "qualification": "The work offers an estimation framework, not a measurement of every proprietary API or deployment stack."
      }
    ],
    "source_ids": [
      "source-2026-10-31-008"
    ],
    "tags": [
      "energy",
      "LLM inference",
      "context windows",
      "sustainability"
    ],
    "image_url": "https://themachinepress.com/issues/2026-10-31/context-energy.webp",
    "corrections": []
  },
  "sources": [
    {
      "source_id": "source-2026-10-31-008",
      "title": "arXiv preprint 2610.39621, original v1",
      "publisher": "arXiv",
      "url": "https://arxiv.org/abs/2610.39621v1",
      "canonical_url": "https://arxiv.org/abs/2610.39621v1",
      "source_type": "primary_research",
      "is_primary_source": true,
      "published_at": "2026-10-28T20:00:00.000-04:00",
      "accessed_at": "2026-10-31T08:25:01.279-04:00",
      "supports_claim_ids": [
        "claim-mp-2026-10-31-008-001"
      ]
    }
  ],
  "corrections": [],
  "publisher": {
    "name": "The Machine Press",
    "url": "https://themachinepress.com",
    "description": "A daily newspaper for the age of artificial intelligence."
  },
  "cite_this_report": {
    "title": "The Input Context Has an Energy Bill Too",
    "publisher": "The Machine Press",
    "published_at": "2026-10-31T09:00:00.000-04:00",
    "canonical_url": "https://themachinepress.com/story/mp-2026-10-31-008/the-input-context-has-an-energy-bill-too"
  }
}