{
  "api_version": "1",
  "updated": "2026-10-29",
  "latest_event": "2026-10-28",
  "rights": "© Frontier Log, all rights reserved.",
  "benchmark": {
    "slug": "agents-last-exam-1-1",
    "name": "Agents' Last Exam v1.1",
    "family": "agents-last-exam",
    "unit": "%",
    "higher_is_better": true,
    "status": "active",
    "active": true,
    "superseded_by": null,
    "retired": null,
    "benchmark_url": "https://agents-last-exam.org",
    "introduced": "2026-10-14",
    "creator": "UC Berkeley RDI",
    "introduced_source": "https://github.com/rdi-berkeley/agents-last-exam/blob/9c5e496e70b52ed104b2060bbf1cbd9bb521c0d9/releases/v1.1/assets.json",
    "settings": [
      {
        "setting": "score",
        "name": "Score",
        "url": "https://frontierlog.ai/benchmark/agents-last-exam-1-1",
        "results": 2
      },
      {
        "setting": "unstated",
        "name": "Measure not stated",
        "url": "https://frontierlog.ai/benchmark/agents-last-exam-1-1/unstated",
        "results": 1
      },
      {
        "setting": "cli-unstated",
        "name": "CLI subset, measure not stated",
        "url": "https://frontierlog.ai/benchmark/agents-last-exam-1-1/cli-unstated",
        "results": 1
      }
    ]
  },
  "total": 4,
  "results": [
    {
      "id": "gemini-4-argon:agents-last-exam-1-1:score:1",
      "model": "gemini-4-argon",
      "model_name": "Gemini 4 Argon",
      "model_date": "2026-09-30",
      "event": "https://frontierlog.ai/event/2026-09-30-gemini-4-argon",
      "benchmark": "agents-last-exam-1-1",
      "benchmark_name": "Agents' Last Exam v1.1",
      "setting": "score",
      "setting_name": "Score",
      "score": 69.0,
      "unit": "%",
      "printed": "69.0%",
      "reported_by": "google",
      "run_by": null,
      "label": "Gemini 4 Argon",
      "source": "https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-4-carbon/",
      "source_title": "Gemini 4 Carbon: our next frontier model",
      "source_date": "2026-10-28",
      "where": "later report: Gemini 4 Carbon post, benchmark table (image), row 'Agents' Last Exam v1.1', 'Score', column 'Gemini 4 Argon'",
      "own": false,
      "primary": true,
      "parallel": false,
      "notes": "122 tasks; ALE-Claw harness; 5-hour window; October 17, 2026 task files; safety filters on",
      "caveat": "",
      "human_verified": false,
      "human_verified_on": null,
      "check": "double_entry",
      "corroborated": false,
      "corroborated_on": null,
      "corroborated_by": [],
      "source_image": "https://frontierlog.ai/sources/bc9b1176d783fc3d.png",
      "source_image_original": "https://storage.googleapis.com/gweb-uniblog-publish-prod/images/gemini_4_carbon__evals-table.original.png"
    },
    {
      "id": "kimi-k3-1-pro:agents-last-exam-1-1:unstated:1",
      "model": "kimi-k3-1-pro",
      "model_name": "Kimi K3.1 Pro",
      "model_date": "2026-10-24",
      "event": "https://frontierlog.ai/event/2026-10-24-kimi-k3-1-pro",
      "benchmark": "agents-last-exam-1-1",
      "benchmark_name": "Agents' Last Exam v1.1",
      "setting": "unstated",
      "setting_name": "Measure not stated",
      "score": 43.2,
      "unit": "%",
      "printed": "43.2",
      "reported_by": "moonshot",
      "run_by": null,
      "label": "Kimi K3.1 Pro (max)",
      "source": "https://www.kimi.ai/blog/kimi-k3-1-pro",
      "source_title": "Kimi K3.1 Pro Tech Blog: Advancing Open Intelligence",
      "source_date": "2026-10-24",
      "where": "'Benchmarks' table (rendered by the page's script), row 'Agents' Last Exam (ALL)', column 'Kimi K3.1 Pro (max)'",
      "own": true,
      "primary": true,
      "parallel": false,
      "notes": "122 tasks; Kimi Code harness; max effort; runs that bypass task requirements scored zero",
      "caveat": "",
      "human_verified": false,
      "human_verified_on": null,
      "check": "double_entry",
      "corroborated": false,
      "corroborated_on": null,
      "corroborated_by": []
    },
    {
      "id": "kimi-k3-1-pro:agents-last-exam-1-1:cli-unstated:1",
      "model": "kimi-k3-1-pro",
      "model_name": "Kimi K3.1 Pro",
      "model_date": "2026-10-24",
      "event": "https://frontierlog.ai/event/2026-10-24-kimi-k3-1-pro",
      "benchmark": "agents-last-exam-1-1",
      "benchmark_name": "Agents' Last Exam v1.1",
      "setting": "cli-unstated",
      "setting_name": "CLI subset, measure not stated",
      "score": 44.2,
      "unit": "%",
      "printed": "44.2",
      "reported_by": "moonshot",
      "run_by": null,
      "label": "Kimi K3.1 Pro (max)",
      "source": "https://www.kimi.ai/blog/kimi-k3-1-pro",
      "source_title": "Kimi K3.1 Pro Tech Blog: Advancing Open Intelligence",
      "source_date": "2026-10-24",
      "where": "'Benchmarks' table (rendered by the page's script), row 'Agents' Last Exam (CLI)', column 'Kimi K3.1 Pro (max)'",
      "own": true,
      "primary": true,
      "parallel": false,
      "notes": "86-task CLI subset; Kimi Code harness; max effort; runs that bypass task requirements scored zero",
      "caveat": "",
      "human_verified": false,
      "human_verified_on": null,
      "check": "double_entry",
      "corroborated": false,
      "corroborated_on": null,
      "corroborated_by": []
    },
    {
      "id": "gemini-4-carbon:agents-last-exam-1-1:score:1",
      "model": "gemini-4-carbon",
      "model_name": "Gemini 4 Carbon",
      "model_date": "2026-10-28",
      "event": "https://frontierlog.ai/event/2026-10-28-gemini-4-carbon",
      "benchmark": "agents-last-exam-1-1",
      "benchmark_name": "Agents' Last Exam v1.1",
      "setting": "score",
      "setting_name": "Score",
      "score": 77.4,
      "unit": "%",
      "printed": "77.4%",
      "reported_by": "google",
      "run_by": null,
      "label": "Gemini 4 Carbon",
      "source": "https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-4-carbon/",
      "source_title": "Gemini 4 Carbon: our next frontier model",
      "source_date": "2026-10-28",
      "where": "benchmark table (image), row 'Agents' Last Exam v1.1', 'Score', column 'Gemini 4 Carbon'",
      "own": true,
      "primary": true,
      "parallel": false,
      "notes": "122 tasks; ALE-Claw harness; 5-hour window; October 17, 2026 task files; safety filters on",
      "caveat": "",
      "human_verified": false,
      "human_verified_on": null,
      "check": "double_entry",
      "corroborated": false,
      "corroborated_on": null,
      "corroborated_by": [],
      "source_image": "https://frontierlog.ai/sources/bc9b1176d783fc3d.png",
      "source_image_original": "https://storage.googleapis.com/gweb-uniblog-publish-prod/images/gemini_4_carbon__evals-table.original.png"
    }
  ]
}