{
  "schemaVersion": "1.2.0",
  "resultId": "builderbench-2026-10-14-wonder-argon",
  "slug": "2026-10-14-wonder-argon",
  "title": "BuilderBench v3: Gemini 4 Argon and Muse Wonder Full",
  "runDate": "2026-10-14",
  "displayDate": "14 October 2026",
  "timezone": "Australia/Perth",
  "status": "cleared_for_web",
  "benchmarkVersion": "v3",
  "packVersion": "3.0.0",
  "packSha256": "37e2da6c5769143b80027b2be176930fd453c5a4b7499001a3692bd8ee664251",
  "methodologyVersion": "v3",
  "scope": "One owner-finalized partial personal review of one Gemini 4 Argon run through the BuilderBench API harness and one Muse Wonder Full run in Meta's Muse Code app on the v3 pack. This is not a universal model ranking or a statistical claim.",
  "durationSeconds": 14400,
  "scoringWeights": {
    "quality": 70,
    "autonomy": 20,
    "proactivity": 10
  },
  "settings": {
    "reasoning": "Extra High requested (Argon ran on High)",
    "speed": "Standard",
    "billing": "Muse Code Power subscription (Muse Wonder Full) and the paid Gemini API (Gemini 4 Argon)"
  },
  "review": {
    "reviewer": "Lennox Saint",
    "sampleSizePerSetup": 1,
    "scoresLocked": true,
    "identitiesRevealed": true,
    "scoreDigestSha256": "f65ab192e77873410aa9bcacfbacbf91f7da99e85ff5c0bef88120bdc872260e",
    "assessment": "Owner-finalized partial personal review",
    "conditions": {
      "mode": "blind",
      "detail": "Requested as a blind comparison. The main build jobs were rated as Candidate A and Candidate B, and the names were revealed after score lock. It was not fully blind: operator logs leaked identity clues, autonomy and proactivity were rated by name, and the native sessions were rated by model name in a separate form."
    }
  },
  "candidates": [
    {
      "name": "Gemini 4 Argon",
      "harness": "BuilderBench API harness",
      "modelVersion": null,
      "harnessVersion": null,
      "quality": 39.12,
      "autonomy": 60,
      "proactivity": 60,
      "overall": 45.38,
      "tasks": [
        {
          "label": "Writing",
          "score": 50,
          "weight": 10,
          "cost": {
            "apiEquivalentEstimateUsd": 1.0488927,
            "coverage": "partial"
          }
        },
        {
          "label": "Landing page",
          "score": 35,
          "weight": 10,
          "cost": {
            "apiEquivalentEstimateUsd": 1.0036189,
            "coverage": "partial"
          }
        },
        {
          "label": "App lead magnet",
          "score": 20,
          "weight": 20,
          "cost": {
            "apiEquivalentEstimateUsd": 2.7256598,
            "coverage": "partial"
          }
        },
        {
          "label": "Video editing",
          "score": 0,
          "weight": 10,
          "cost": {
            "apiEquivalentEstimateUsd": null,
            "coverage": "unavailable"
          }
        },
        {
          "label": "Motion",
          "score": 25,
          "weight": 5,
          "cost": {
            "apiEquivalentEstimateUsd": null,
            "coverage": "unavailable"
          }
        },
        {
          "label": "Thumbnails",
          "score": 62,
          "weight": 10,
          "cost": {
            "apiEquivalentEstimateUsd": 3.0036537,
            "coverage": "partial"
          }
        },
        {
          "label": "Browser + computer",
          "score": 20,
          "weight": 10,
          "cost": {
            "apiEquivalentEstimateUsd": null,
            "coverage": "unavailable"
          }
        },
        {
          "label": "Research",
          "score": 45,
          "weight": 10,
          "cost": {
            "apiEquivalentEstimateUsd": 0.3998395,
            "coverage": "partial"
          }
        },
        {
          "label": "Automation",
          "score": 91.67,
          "weight": 10,
          "cost": {
            "apiEquivalentEstimateUsd": 0.7423089,
            "coverage": "partial"
          }
        },
        {
          "label": "Founder conversations",
          "score": 70,
          "weight": 5,
          "cost": {
            "apiEquivalentEstimateUsd": null,
            "coverage": "unavailable"
          }
        }
      ],
      "cost": {
        "actualIncrementalSpendUsd": null,
        "apiEquivalentEstimateUsd": 8.9239735,
        "coverage": "Partial API-equivalent estimate: 15 of 31 saved execution records priced from provider-reported estimates. Unpriced records, native sessions and failed attempts are not included.",
        "pricedRecords": {
          "priced": 15,
          "total": 31
        },
        "estimate": {
          "status": "partial",
          "method": "Provider-reported estimates for recorded execution usage.",
          "pricedAt": "2026-10-14"
        }
      },
      "apiPricing": {
        "status": "verified",
        "unit": "per 1M tokens",
        "processing": "Standard",
        "checkedAt": "2026-10-14",
        "sourceUrl": "https://ai.google.dev/gemini-api/docs/latest-model",
        "sourceLabel": "Google Gemini API docs: Gemini 4 Argon",
        "tiers": [
          {
            "label": "Introductory, through 30 Nov 2026",
            "inputUsd": 2,
            "cacheReadUsd": 0.1,
            "outputUsd": 10
          }
        ],
        "notes": "From 1 December 2026: $4.00 input, $0.20 cache read and $20.00 output. Output includes thinking tokens. Cache reads are implicit; explicit cache writes are unsupported. 1M context with no separate long-context price. API use is billed separately from Google subscriptions."
      },
      "nativeRatings": [
        {
          "lane": "browser",
          "score": 20
        },
        {
          "lane": "computer",
          "score": 20
        },
        {
          "lane": "advice",
          "score": 70
        },
        {
          "lane": "planning",
          "score": 70
        }
      ]
    },
    {
      "name": "Muse Wonder Full",
      "harness": "Muse Code",
      "modelVersion": null,
      "harnessVersion": null,
      "quality": 37.73,
      "autonomy": 45,
      "proactivity": 45,
      "overall": 39.91,
      "tasks": [
        {
          "label": "Writing",
          "score": 40,
          "weight": 10,
          "cost": {
            "apiEquivalentEstimateUsd": null,
            "coverage": "unavailable"
          }
        },
        {
          "label": "Landing page",
          "score": 60,
          "weight": 10,
          "cost": {
            "apiEquivalentEstimateUsd": null,
            "coverage": "unavailable"
          }
        },
        {
          "label": "App lead magnet",
          "score": 20,
          "weight": 20,
          "cost": {
            "apiEquivalentEstimateUsd": null,
            "coverage": "unavailable"
          }
        },
        {
          "label": "Video editing",
          "score": 33.33,
          "weight": 10,
          "cost": {
            "apiEquivalentEstimateUsd": null,
            "coverage": "unavailable"
          }
        },
        {
          "label": "Motion",
          "score": 30,
          "weight": 5,
          "cost": {
            "apiEquivalentEstimateUsd": null,
            "coverage": "unavailable"
          }
        },
        {
          "label": "Thumbnails",
          "score": 39,
          "weight": 10,
          "cost": {
            "apiEquivalentEstimateUsd": null,
            "coverage": "unavailable"
          }
        },
        {
          "label": "Browser + computer",
          "score": 0,
          "weight": 10,
          "cost": {
            "apiEquivalentEstimateUsd": null,
            "coverage": "unavailable"
          }
        },
        {
          "label": "Research",
          "score": 50,
          "weight": 10,
          "cost": {
            "apiEquivalentEstimateUsd": null,
            "coverage": "unavailable"
          }
        },
        {
          "label": "Automation",
          "score": 75,
          "weight": 10,
          "cost": {
            "apiEquivalentEstimateUsd": null,
            "coverage": "unavailable"
          }
        },
        {
          "label": "Founder conversations",
          "score": 50,
          "weight": 5,
          "cost": {
            "apiEquivalentEstimateUsd": null,
            "coverage": "unavailable"
          }
        }
      ],
      "cost": {
        "actualIncrementalSpendUsd": null,
        "apiEquivalentEstimateUsd": null,
        "coverage": "Unavailable: Muse Code reported no token usage or cost for these runs, so the run cost is unknown, not zero.",
        "pricedRecords": {
          "priced": 0,
          "total": 23
        },
        "estimate": {
          "status": "unavailable",
          "method": "None: no supported cost basis for a subscription-only model.",
          "pricedAt": null
        }
      },
      "apiPricing": {
        "status": "unavailable",
        "reason": "not_published",
        "detail": "Meta offers Muse Wonder Full only inside Muse Code subscriptions (Power plan, $50 a month) and publishes no API token prices. A plan price is not a run bill, and regular Muse Wonder rates are not substituted.",
        "checkedAt": "2026-10-13"
      },
      "nativeRatings": [
        {
          "lane": "browser",
          "score": 0
        },
        {
          "lane": "computer",
          "score": 0
        },
        {
          "lane": "advice",
          "score": 40
        },
        {
          "lane": "planning",
          "score": 60
        }
      ]
    }
  ],
  "costNote": "API-equivalent estimates price recorded execution usage only. They are not bills. Gemini 4 Argon keeps provider-reported estimates for 15 of 31 saved records, so its figure is partial. Muse Wonder Full ran on Meta's Muse Code Power plan, which reported no usage for these runs, so its run cost is unknown, not zero. The $50-a-month plan price is not a run bill. Native sessions, setup, calibration, coordination, grading, presentation and subscription fees are excluded. Actual incremental spend is unknown.",
  "methodologyNote": "The owner rated the saved build outputs, the four native sessions, autonomy and proactivity. The automation repair task was conductor-scored from recorded evidence. The saved scores were finalized as a partial personal review and locked before the names were revealed. Quality is the weighted task score. v1 to v3 use 70% quality, 20% autonomy and 10% proactivity.",
  "limitations": [
    "This is one owner-finalized partial personal review of one run per setup. It is not a universal ranking and it is not statistically conclusive.",
    "Finalized ratings do not mean every task or recording passed verification. The original execution and its evidence are incomplete.",
    "A 5.47-point gap from one run per setup does not show that one model is better in general.",
    "Extra High effort was requested. Gemini 4 Argon's highest available setting was High, so it ran on High.",
    "The two setups used different tools: Muse Wonder Full worked inside Meta's Muse Code app, and Gemini 4 Argon worked through the BuilderBench API harness.",
    "A storage problem interrupted the four-hour build window, and the clock was not reset.",
    "Three of the four native session recordings have coverage gaps, and the fourth has not been accepted.",
    "The review was not fully blind. Operator logs leaked identity clues before the reveal, and autonomy and proactivity were rated by name.",
    "The native founder conversations were captured separately from the build window.",
    "Video, motion and thumbnail files passed technical decoding checks only. Full behavioral, responsive and audiovisual review is incomplete.",
    "Muse Wonder Full's run cost is unknown. Its $50-a-month plan is a subscription price, not a bill for this run.",
    "Gemini 4 Argon's cost is a partial API-equivalent estimate, not a bill. Actual incremental spend is unknown.",
    "The score belongs to the full setup: model, harness, tools and settings.",
    "Pack 3.0.0 scores compare only with other pack 3.0.0 results. Opus 5.5's 68.26 comes from the pack 1.0.0 pilot and is a dated reference, not a same-pack comparison.",
    "The build window was four hours. Native live sessions were timed separately from the build window."
  ],
  "assets": {
    "resultGraphic": {
      "src": "/builderbench/2026-10-14-wonder-argon/wonder-argon-reveal.png",
      "sha256": "fce369ad1d1e73b36062e7aec155e9bc173ec2766f8843ee76db63d8163e611b",
      "width": 1672,
      "height": 941,
      "alt": "BuilderBench by the LAB, the reveal: Gemini 4 Argon scored 45.38 out of 100 and Muse Wonder Full scored 39.91 out of 100. Argon leads by 5.47 points. Owner-finalized partial review with partial evidence, 14 October 2026, v3, four hours, Extra High requested and Argon ran on High. Argon's API estimate is US$8.92 from partial recorded usage, not a bill. Muse Wonder Full's run cost is unknown. Argon base rates per 1M tokens: $2.00 input, $0.10 cache read, $10.00 output until 30 November, then $4.00, $0.20 and $20.00. Muse Wonder Full has no published API price."
    }
  },
  "sourceResultsSha256": "3616aa2977f25869d5419cec0faff7e2a0be3fea965a15969cac0088d242a5f6",
  "integrity": {
    "algorithm": "sha256",
    "digest": "a311e1ca284090829bed54fcaaca902028a1d9597896e035cee4958336f4ebc7"
  }
}