{
  "api_version": "1",
  "updated": "2026-10-26",
  "latest_event": "2026-10-24",
  "rights": "© Frontier Log, all rights reserved.",
  "model": {
    "slug": "glm-5-1",
    "name": "GLM-5.1",
    "lab": "zai",
    "date": "2026-04-07",
    "flagship": false,
    "event": "https://frontierlog.ai/event/2026-04-07-glm-5-1",
    "releases": [],
    "variant_of": null,
    "made_by": "Z.ai",
    "url": "https://frontierlog.ai/model/glm-5-1",
    "results": [
      {
        "id": "glm-5-1:hle:no-tools:1",
        "model": "glm-5-1",
        "model_name": "GLM-5.1",
        "model_date": "2026-04-07",
        "event": "https://frontierlog.ai/event/2026-04-07-glm-5-1",
        "benchmark": "hle",
        "benchmark_name": "Humanity's Last Exam",
        "setting": "no-tools",
        "setting_name": "No tools",
        "score": 31.0,
        "unit": "%",
        "printed": "31.0",
        "reported_by": "zai",
        "run_by": null,
        "label": "GLM-5.1",
        "source": "https://z.ai/blog/glm-5.1",
        "source_title": "GLM-5.1: Towards Long-Horizon Tasks",
        "source_date": "2026-04-07",
        "where": "benchmark table near the end, row 'HLE', column 'GLM-5.1'",
        "own": true,
        "primary": true,
        "parallel": false,
        "notes": "text-only subset; GPT-5.2 judge; up to 163,840 output tokens",
        "caveat": "",
        "human_verified": true,
        "human_verified_on": "2026-10-19",
        "check": "human",
        "corroborated": true,
        "corroborated_on": "2026-10-19",
        "corroborated_by": [
          {
            "url": "https://z.ai/blog/glm-5.2",
            "name": "GLM-5.2",
            "title": "GLM-5.2: Built for Long-Horizon Tasks",
            "lab": "zai",
            "date": "2026-06-16",
            "printed": "31.0",
            "where": "Chart 'LLM Performance Evaluation' (all models at max thinking effort), panel 'Humanity's Last Exam', bar 'GLM-5.1' (lower label, not marked w/ Tools)"
          }
        ],
        "source_capture": "https://frontierlog.ai/captures/35c66845b36400f8.png"
      },
      {
        "id": "glm-5-1:hle:with-tools:1",
        "model": "glm-5-1",
        "model_name": "GLM-5.1",
        "model_date": "2026-04-07",
        "event": "https://frontierlog.ai/event/2026-04-07-glm-5-1",
        "benchmark": "hle",
        "benchmark_name": "Humanity's Last Exam",
        "setting": "with-tools",
        "setting_name": "With tools",
        "score": 52.3,
        "unit": "%",
        "printed": "52.3",
        "reported_by": "zai",
        "run_by": null,
        "label": "GLM-5.1",
        "source": "https://z.ai/blog/glm-5.1",
        "source_title": "GLM-5.1: Towards Long-Horizon Tasks",
        "source_date": "2026-04-07",
        "where": "benchmark table near the end, row 'HLE w/ Tools', column 'GLM-5.1'",
        "own": true,
        "primary": true,
        "parallel": false,
        "notes": "text-only subset; GPT-5.2 judge; 202,752-token context",
        "caveat": "",
        "human_verified": true,
        "human_verified_on": "2026-10-19",
        "check": "human",
        "corroborated": true,
        "corroborated_on": "2026-10-19",
        "corroborated_by": [
          {
            "url": "https://z.ai/blog/glm-5.2",
            "name": "GLM-5.2",
            "title": "GLM-5.2: Built for Long-Horizon Tasks",
            "lab": "zai",
            "date": "2026-06-16",
            "printed": "52.3",
            "where": "Full Benchmark Table, row 'HLE w/ Tools (* = full set for Opus 4.8, GPT-5.5, Gemini 3.1 Pro)', column 'GLM-5.1'"
          }
        ],
        "source_capture": "https://frontierlog.ai/captures/35c66845b36400f8.png"
      },
      {
        "id": "glm-5-1:gpqa-diamond:default:1",
        "model": "glm-5-1",
        "model_name": "GLM-5.1",
        "model_date": "2026-04-07",
        "event": "https://frontierlog.ai/event/2026-04-07-glm-5-1",
        "benchmark": "gpqa-diamond",
        "benchmark_name": "GPQA Diamond",
        "setting": "default",
        "setting_name": "",
        "score": 86.2,
        "unit": "%",
        "printed": "86.2",
        "reported_by": "zai",
        "run_by": null,
        "label": "GLM-5.1",
        "source": "https://z.ai/blog/glm-5.1",
        "source_title": "GLM-5.1: Towards Long-Horizon Tasks",
        "source_date": "2026-04-07",
        "where": "benchmark table near the end, row 'GPQA-Diamond', column 'GLM-5.1'",
        "own": true,
        "primary": true,
        "parallel": false,
        "notes": "up to 163,840 output tokens",
        "caveat": "",
        "human_verified": true,
        "human_verified_on": "2026-10-19",
        "check": "human",
        "corroborated": true,
        "corroborated_on": "2026-10-19",
        "corroborated_by": [
          {
            "url": "https://api-docs.deepseek.com/news/news260424/",
            "name": "DeepSeek V4 Preview",
            "title": "DeepSeek V4 Preview Release",
            "lab": "deepseek",
            "date": "2026-04-24",
            "printed": "86.2",
            "where": "Full benchmark table image; row 'GPQA Diamond (Pass@1)'; column 'GLM-5.1 Thinking'"
          },
          {
            "url": "https://z.ai/blog/glm-5.2",
            "name": "GLM-5.2",
            "title": "GLM-5.2: Built for Long-Horizon Tasks",
            "lab": "zai",
            "date": "2026-06-16",
            "printed": "86.2",
            "where": "Full Benchmark Table, row 'GPQA-Diamond', column 'GLM-5.1'"
          }
        ],
        "source_capture": "https://frontierlog.ai/captures/35c66845b36400f8.png"
      },
      {
        "id": "glm-5-1:swe-bench-pro:default:1",
        "model": "glm-5-1",
        "model_name": "GLM-5.1",
        "model_date": "2026-04-07",
        "event": "https://frontierlog.ai/event/2026-04-07-glm-5-1",
        "benchmark": "swe-bench-pro",
        "benchmark_name": "SWE-bench Pro",
        "setting": "default",
        "setting_name": "",
        "score": 58.4,
        "unit": "%",
        "printed": "58.4",
        "reported_by": "zai",
        "run_by": null,
        "label": "GLM-5.1",
        "source": "https://z.ai/blog/glm-5.1",
        "source_title": "GLM-5.1: Towards Long-Horizon Tasks",
        "source_date": "2026-04-07",
        "where": "benchmark table near the end, row 'SWE-Bench Pro', column 'GLM-5.1'",
        "own": true,
        "primary": true,
        "parallel": false,
        "notes": "OpenHands with a tailored prompt; 200K context",
        "caveat": "The source doesn't say whether this is the public or the commercial set",
        "human_verified": true,
        "human_verified_on": "2026-10-19",
        "check": "human",
        "corroborated": true,
        "corroborated_on": "2026-10-19",
        "corroborated_by": [
          {
            "url": "https://api-docs.deepseek.com/news/news260424/",
            "name": "DeepSeek V4 Preview",
            "title": "DeepSeek V4 Preview Release",
            "lab": "deepseek",
            "date": "2026-04-24",
            "printed": "58.4",
            "where": "Full benchmark table image; row 'SWE Pro (Resolved)'; column 'GLM-5.1 Thinking'"
          },
          {
            "url": "https://www.minimax.io/blog/minimax-m3",
            "name": "MiniMax M3",
            "title": "MiniMax M3: Frontier Coding, 1M Context, Native Multimodality — All in One Model",
            "lab": "minimax",
            "date": "2026-06-01",
            "printed": "58.4",
            "where": "Full benchmark table image; row 'SWE-Bench Pro'; column 'GLM 5.1 Thinking'"
          },
          {
            "url": "https://qwen.ai/blog?id=qwen3.6-max-preview",
            "name": "Qwen3.6-Max-Preview",
            "title": "Qwen3.6-Max-Preview: Smarter, Sharper, Still Evolving",
            "lab": "alibaba",
            "date": "2026-04-18",
            "printed": "58.4",
            "where": "Score chart image (qwen3.6_max_preview_score.png), panel SWE-bench Pro, value label on the GLM 5.1 bar"
          },
          {
            "url": "https://z.ai/blog/glm-5.2",
            "name": "GLM-5.2",
            "title": "GLM-5.2: Built for Long-Horizon Tasks",
            "lab": "zai",
            "date": "2026-06-16",
            "printed": "58.4",
            "where": "Prose: '62.1 vs. 58.4 on SWE-bench Pro' (GLM-5.1)"
          }
        ],
        "source_capture": "https://frontierlog.ai/captures/35c66845b36400f8.png"
      },
      {
        "id": "glm-5-1:terminal-bench-2-0:default:1",
        "model": "glm-5-1",
        "model_name": "GLM-5.1",
        "model_date": "2026-04-07",
        "event": "https://frontierlog.ai/event/2026-04-07-glm-5-1",
        "benchmark": "terminal-bench-2-0",
        "benchmark_name": "Terminal-Bench 2.0",
        "setting": "default",
        "setting_name": "",
        "score": 63.5,
        "unit": "%",
        "printed": "63.5",
        "reported_by": "zai",
        "run_by": null,
        "label": "GLM-5.1",
        "source": "https://z.ai/blog/glm-5.1",
        "source_title": "GLM-5.1: Towards Long-Horizon Tasks",
        "source_date": "2026-04-07",
        "where": "benchmark table near the end, row 'Terminal-Bench 2.0 Terminus-2', column 'GLM-5.1'",
        "own": true,
        "primary": true,
        "parallel": false,
        "notes": "Terminus 2 harness; 3h timeout; 200K context",
        "caveat": "",
        "human_verified": true,
        "human_verified_on": "2026-10-19",
        "check": "human",
        "corroborated": true,
        "corroborated_on": "2026-10-19",
        "corroborated_by": [
          {
            "url": "https://api-docs.deepseek.com/news/news260424/",
            "name": "DeepSeek V4 Preview",
            "title": "DeepSeek V4 Preview Release",
            "lab": "deepseek",
            "date": "2026-04-24",
            "printed": "63.5",
            "where": "Full benchmark table image; row 'Terminal Bench 2.0 (Acc)'; column 'GLM-5.1 Thinking'"
          },
          {
            "url": "https://qwen.ai/blog?id=qwen3.6-max-preview",
            "name": "Qwen3.6-Max-Preview",
            "title": "Qwen3.6-Max-Preview: Smarter, Sharper, Still Evolving",
            "lab": "alibaba",
            "date": "2026-04-18",
            "printed": "63.5",
            "where": "Score chart image (qwen3.6_max_preview_score.png), panel Terminal-Bench 2.0 (Terminus-2), value label on the GLM 5.1 bar"
          }
        ],
        "source_capture": "https://frontierlog.ai/captures/35c66845b36400f8.png"
      },
      {
        "id": "glm-5-1:terminal-bench-2-0:default:2",
        "model": "glm-5-1",
        "model_name": "GLM-5.1",
        "model_date": "2026-04-07",
        "event": "https://frontierlog.ai/event/2026-04-07-glm-5-1",
        "benchmark": "terminal-bench-2-0",
        "benchmark_name": "Terminal-Bench 2.0",
        "setting": "default",
        "setting_name": "",
        "score": 69.0,
        "unit": "%",
        "printed": "69.0",
        "reported_by": "zai",
        "run_by": null,
        "label": "GLM-5.1",
        "source": "https://z.ai/blog/glm-5.1",
        "source_title": "GLM-5.1: Towards Long-Horizon Tasks",
        "source_date": "2026-04-07",
        "where": "benchmark table near the end, row 'Terminal-Bench 2.0 Best self-reported harness', column 'GLM-5.1'",
        "own": true,
        "primary": false,
        "parallel": false,
        "notes": "Claude Code harness, thinking; no time limit; averaged over 5 runs",
        "caveat": "",
        "human_verified": true,
        "human_verified_on": "2026-10-19",
        "check": "human",
        "corroborated": false,
        "corroborated_on": null,
        "corroborated_by": [],
        "source_capture": "https://frontierlog.ai/captures/35c66845b36400f8.png"
      },
      {
        "id": "glm-5-1:browsecomp:default:1",
        "model": "glm-5-1",
        "model_name": "GLM-5.1",
        "model_date": "2026-04-07",
        "event": "https://frontierlog.ai/event/2026-04-07-glm-5-1",
        "benchmark": "browsecomp",
        "benchmark_name": "BrowseComp",
        "setting": "default",
        "setting_name": "",
        "score": 68.0,
        "unit": "%",
        "printed": "68.0",
        "reported_by": "zai",
        "run_by": null,
        "label": "GLM-5.1",
        "source": "https://z.ai/blog/glm-5.1",
        "source_title": "GLM-5.1: Towards Long-Horizon Tasks",
        "source_date": "2026-04-07",
        "where": "benchmark table near the end, row 'BrowseComp', column 'GLM-5.1'",
        "own": true,
        "primary": true,
        "parallel": false,
        "notes": "without context management: keeps the last 5 turns",
        "caveat": "",
        "human_verified": true,
        "human_verified_on": "2026-10-19",
        "check": "human",
        "corroborated": false,
        "corroborated_on": null,
        "corroborated_by": [],
        "source_capture": "https://frontierlog.ai/captures/35c66845b36400f8.png"
      },
      {
        "id": "glm-5-1:browsecomp:default:2",
        "model": "glm-5-1",
        "model_name": "GLM-5.1",
        "model_date": "2026-04-07",
        "event": "https://frontierlog.ai/event/2026-04-07-glm-5-1",
        "benchmark": "browsecomp",
        "benchmark_name": "BrowseComp",
        "setting": "default",
        "setting_name": "",
        "score": 79.3,
        "unit": "%",
        "printed": "79.3",
        "reported_by": "zai",
        "run_by": null,
        "label": "GLM-5.1",
        "source": "https://z.ai/blog/glm-5.1",
        "source_title": "GLM-5.1: Towards Long-Horizon Tasks",
        "source_date": "2026-04-07",
        "where": "benchmark table near the end, row 'BrowseComp w/ Context Manage', column 'GLM-5.1'",
        "own": true,
        "primary": false,
        "parallel": false,
        "notes": "with context management, discard-all strategy",
        "caveat": "",
        "human_verified": true,
        "human_verified_on": "2026-10-19",
        "check": "human",
        "corroborated": true,
        "corroborated_on": "2026-10-19",
        "corroborated_by": [
          {
            "url": "https://api-docs.deepseek.com/news/news260424/",
            "name": "DeepSeek V4 Preview",
            "title": "DeepSeek V4 Preview Release",
            "lab": "deepseek",
            "date": "2026-04-24",
            "printed": "79.3",
            "where": "Full benchmark table image; row 'BrowseComp (Pass@1)'; column 'GLM-5.1 Thinking'"
          },
          {
            "url": "https://www.minimax.io/blog/minimax-m3",
            "name": "MiniMax M3",
            "title": "MiniMax M3: Frontier Coding, 1M Context, Native Multimodality — All in One Model",
            "lab": "minimax",
            "date": "2026-06-01",
            "printed": "79.3",
            "where": "Full benchmark table image; row 'BrowseComp'; column 'GLM 5.1 Thinking'"
          },
          {
            "url": "https://mistral.ai/news/vibe-remote-agents-mistral-medium-3-5/",
            "name": "Mistral Medium 3.5",
            "title": "Remote agents in Vibe. Powered by Mistral Medium 3.5.",
            "lab": "mistral",
            "date": "2026-05-22",
            "printed": "79.3",
            "where": "Chart 'Agentic Benchmarks vs competing models', group 'BrowseComp****' (footnote: self-reported, Mistral uses context management, discard-all at 100k), bar 'GLM-5.1 744B A40B'"
          }
        ],
        "source_capture": "https://frontierlog.ai/captures/35c66845b36400f8.png"
      },
      {
        "id": "glm-5-1:mcp-atlas:default:1",
        "model": "glm-5-1",
        "model_name": "GLM-5.1",
        "model_date": "2026-04-07",
        "event": "https://frontierlog.ai/event/2026-04-07-glm-5-1",
        "benchmark": "mcp-atlas",
        "benchmark_name": "MCP Atlas",
        "setting": "default",
        "setting_name": "",
        "score": 71.8,
        "unit": "%",
        "printed": "71.8",
        "reported_by": "zai",
        "run_by": null,
        "label": "GLM-5.1",
        "source": "https://z.ai/blog/glm-5.1",
        "source_title": "GLM-5.1: Towards Long-Horizon Tasks",
        "source_date": "2026-04-07",
        "where": "benchmark table near the end, row 'MCP-Atlas Public Set', column 'GLM-5.1'",
        "own": true,
        "primary": true,
        "parallel": false,
        "notes": "500-task public set; thinking mode; 10-minute timeout per task; Gemini 3 Pro judge",
        "caveat": "",
        "human_verified": true,
        "human_verified_on": "2026-10-19",
        "check": "human",
        "corroborated": true,
        "corroborated_on": "2026-10-19",
        "corroborated_by": [
          {
            "url": "https://api-docs.deepseek.com/news/news260424/",
            "name": "DeepSeek V4 Preview",
            "title": "DeepSeek V4 Preview Release",
            "lab": "deepseek",
            "date": "2026-04-24",
            "printed": "71.8",
            "where": "Full benchmark table image; row 'MCPAtlas Public (Pass@1)'; column 'GLM-5.1 Thinking'"
          },
          {
            "url": "https://www.minimax.io/blog/minimax-m3",
            "name": "MiniMax M3",
            "title": "MiniMax M3: Frontier Coding, 1M Context, Native Multimodality — All in One Model",
            "lab": "minimax",
            "date": "2026-06-01",
            "printed": "71.8",
            "where": "Full benchmark table image; row 'MCP Atlas'; column 'GLM 5.1 Thinking'"
          },
          {
            "url": "https://qwen.ai/blog?id=qwen3.7",
            "name": "Qwen3.7-Max",
            "title": "Qwen3.7: The Agent Frontier",
            "lab": "alibaba",
            "date": "2026-05-20",
            "printed": "71.8",
            "where": "Score chart image (Qwen3.7-Max-Score.png), panel MCP-Atlas, value label on the GLM-5.1 bar"
          },
          {
            "url": "https://z.ai/blog/glm-5.2",
            "name": "GLM-5.2",
            "title": "GLM-5.2: Built for Long-Horizon Tasks",
            "lab": "zai",
            "date": "2026-06-16",
            "printed": "71.8",
            "where": "Full Benchmark Table, row 'MCP-Atlas Public Set', column 'GLM-5.1'"
          }
        ],
        "source_capture": "https://frontierlog.ai/captures/35c66845b36400f8.png"
      },
      {
        "id": "glm-5-1:deepswe-1-1:default:1",
        "model": "glm-5-1",
        "model_name": "GLM-5.1",
        "model_date": "2026-04-07",
        "event": "https://frontierlog.ai/event/2026-04-07-glm-5-1",
        "benchmark": "deepswe-1-1",
        "benchmark_name": "DeepSWE v1.1",
        "setting": "default",
        "setting_name": "",
        "score": 18.0,
        "unit": "%",
        "printed": "18.0",
        "reported_by": "zai",
        "run_by": null,
        "label": "GLM-5.1",
        "source": "https://z.ai/blog/glm-5.2",
        "source_title": "GLM-5.2: Built for Long-Horizon Tasks",
        "source_date": "2026-06-16",
        "where": "'Full Benchmark Table', row 'DeepSWE', column 'GLM-5.1'",
        "own": false,
        "primary": true,
        "parallel": false,
        "notes": "mini-swe-agent harness",
        "caveat": "",
        "human_verified": true,
        "human_verified_on": "2026-10-19",
        "check": "human",
        "corroborated": false,
        "corroborated_on": null,
        "corroborated_by": [],
        "source_capture": "https://frontierlog.ai/captures/112d3aba81edfc46.png"
      },
      {
        "id": "glm-5-1:terminal-bench-2-1:default:1",
        "model": "glm-5-1",
        "model_name": "GLM-5.1",
        "model_date": "2026-04-07",
        "event": "https://frontierlog.ai/event/2026-04-07-glm-5-1",
        "benchmark": "terminal-bench-2-1",
        "benchmark_name": "Terminal-Bench 2.1",
        "setting": "default",
        "setting_name": "",
        "score": 63.5,
        "unit": "%",
        "printed": "63.5",
        "reported_by": "zai",
        "run_by": null,
        "label": "GLM-5.1",
        "source": "https://z.ai/blog/glm-5.2",
        "source_title": "GLM-5.2: Built for Long-Horizon Tasks",
        "source_date": "2026-06-16",
        "where": "'Full Benchmark Table', row 'Terminal Bench 2.1 Terminus-2', column 'GLM-5.1'",
        "own": false,
        "primary": true,
        "parallel": false,
        "notes": "Terminus 2 harness",
        "caveat": "",
        "human_verified": true,
        "human_verified_on": "2026-10-19",
        "check": "human",
        "corroborated": false,
        "corroborated_on": null,
        "corroborated_by": [],
        "source_capture": "https://frontierlog.ai/captures/112d3aba81edfc46.png"
      },
      {
        "id": "glm-5-1:terminal-bench-2-1:default:2",
        "model": "glm-5-1",
        "model_name": "GLM-5.1",
        "model_date": "2026-04-07",
        "event": "https://frontierlog.ai/event/2026-04-07-glm-5-1",
        "benchmark": "terminal-bench-2-1",
        "benchmark_name": "Terminal-Bench 2.1",
        "setting": "default",
        "setting_name": "",
        "score": 69.0,
        "unit": "%",
        "printed": "69",
        "reported_by": "zai",
        "run_by": null,
        "label": "GLM-5.1",
        "source": "https://z.ai/blog/glm-5.2",
        "source_title": "GLM-5.2: Built for Long-Horizon Tasks",
        "source_date": "2026-06-16",
        "where": "'Full Benchmark Table', row 'Terminal Bench 2.1 Best Reported Harness', column 'GLM-5.1'",
        "own": false,
        "primary": false,
        "parallel": false,
        "notes": "Claude Code harness",
        "caveat": "",
        "human_verified": true,
        "human_verified_on": "2026-10-19",
        "check": "human",
        "corroborated": false,
        "corroborated_on": null,
        "corroborated_by": [],
        "source_capture": "https://frontierlog.ai/captures/112d3aba81edfc46.png"
      }
    ]
  }
}