data: update benchmark scores 2026-07-17

Author: AGI Clock Bot

src/data/live.json

@@ -1,14 +1,14 @@
 {
   "schemaVersion": 2,
   "formula": "agi-readiness-v2",
-  "updatedAt": "2026-07-16T05:14:47.391Z",
-  "agiReadiness": 75.6,
-  "agiDistance": 24.4,
+  "updatedAt": "2026-07-17T05:17:41.505Z",
+  "agiReadiness": 76.1,
+  "agiDistance": 23.9,
   "bottleneckDimension": {
     "id": "agentsAutonomy",
     "name": "Agents + Autonomy",
-    "score": 59.1,
-    "distance": 40.9
+    "score": 59.8,
+    "distance": 40.2
   },
   "eta": {
     "status": "collecting_baseline",
@@ -40,17 +40,17 @@
       "name": "Agents + Autonomy",
       "weight": 0.3,
       "description": "Computer use, long-horizon professional workflows, planning, and applied office/finance tasks.",
-      "score": 59.1,
-      "distance": 40.9,
+      "score": 59.8,
+      "distance": 40.2,
       "benchmarkCount": 6
     },
     {
       "id": "toolSearchReliability",
       "name": "Tool Use + Search Reliability",
       "weight": 0.15,
       "description": "Tool calling, web/search persistence, user-policy following, and repeated agent reliability proxies.",
-      "score": 81.6,
-      "distance": 18.4,
+      "score": 83.6,
+      "distance": 16.4,
       "benchmarkCount": 8
     }
   ],
@@ -80,7 +80,7 @@
       "scoreRaw": 0.647,
       "modelName": "Claude Mythos Preview",
       "modelId": "claude-mythos-preview",
-      "totalModels": 88,
+      "totalModels": 89,
       "trend": "flat",
       "source": "llm-stats:humanity's-last-exam",
       "description": "2,500 expert-vetted multimodal questions across math, science, humanities, and vision."
@@ -95,7 +95,7 @@
       "scoreRaw": 0.946,
       "modelName": "GPT-5.6 Sol",
       "modelId": "gpt-5.6-sol",
-      "totalModels": 231,
+      "totalModels": 232,
       "trend": "flat",
       "source": "llm-stats:gpqa",
       "description": "PhD-level questions in biology, chemistry, and physics."
@@ -126,7 +126,7 @@
       "modelName": "GPT-5.6 Sol",
       "modelId": "gpt-5.6-sol",
       "totalModels": 17,
-      "trend": "up",
+      "trend": "flat",
       "source": "llm-stats:frontiermath",
       "description": "Advanced unpublished mathematical problems authored and reviewed by expert mathematicians."
     },
@@ -200,7 +200,7 @@
       "scoreRaw": 0.888,
       "modelName": "GPT-5.6 Sol",
       "modelId": "gpt-5.6-sol",
-      "totalModels": 11,
+      "totalModels": 12,
       "trend": "flat",
       "source": "llm-stats:terminal-bench-2.1",
       "description": "Updated terminal-agent benchmark for autonomous command-line execution."
@@ -226,12 +226,12 @@
       "name": "APEX-Agents",
       "category": "Long-Horizon Agents",
       "dimension": "agentsAutonomy",
-      "score": 33.8,
-      "scoreRaw": 0.338,
-      "modelName": "Seed 2.1 Pro",
-      "modelId": "seed-2.1-pro",
-      "totalModels": 6,
-      "trend": "flat",
+      "score": 37.6,
+      "scoreRaw": 0.376,
+      "modelName": "Kimi K3",
+      "modelId": "kimi-k3",
+      "totalModels": 7,
+      "trend": "up",
       "source": "llm-stats:apex-agents",
       "description": "Long-horizon professional tasks requiring sustained planning and execution."
     },
@@ -275,7 +275,7 @@
       "scoreRaw": 0.722,
       "modelName": "Seed 2.1 Pro",
       "modelId": "seed-2.1-pro",
-      "totalModels": 6,
+      "totalModels": 7,
       "trend": "flat",
       "source": "llm-stats:officeqa-pro",
       "description": "Professional knowledge-work tasks involving documents, spreadsheets, and office workflows."
@@ -346,12 +346,12 @@
       "name": "MCP Atlas",
       "category": "Tool Use",
       "dimension": "toolSearchReliability",
-      "score": 83.8,
-      "scoreRaw": 0.838,
-      "modelName": "Seed 2.1 Pro",
-      "modelId": "seed-2.1-pro",
-      "totalModels": 28,
-      "trend": "flat",
+      "score": 84.2,
+      "scoreRaw": 0.842,
+      "modelName": "Kimi K3",
+      "modelId": "kimi-k3",
+      "totalModels": 29,
+      "trend": "up",
       "source": "llm-stats:mcp-atlas",
       "description": "Scaled tool coordination across complex multi-step tool-use tasks."
     },
@@ -361,12 +361,12 @@
       "name": "Toolathlon",
       "category": "Tool Use",
       "dimension": "toolSearchReliability",
-      "score": 59.9,
-      "scoreRaw": 0.599,
-      "modelName": "Claude Opus 4.8",
-      "modelId": "claude-opus-4-8",
-      "totalModels": 28,
-      "trend": "flat",
+      "score": 73.2,
+      "scoreRaw": 0.732,
+      "modelName": "Kimi K3",
+      "modelId": "kimi-k3",
+      "totalModels": 29,
+      "trend": "up",
       "source": "llm-stats:toolathlon",
       "description": "Multi-tool proficiency across diverse tool-use categories."
     },
@@ -376,12 +376,12 @@
       "name": "BrowseComp",
       "category": "Search",
       "dimension": "toolSearchReliability",
-      "score": 90.4,
-      "scoreRaw": 0.904,
-      "modelName": "GPT-5.6 Sol",
-      "modelId": "gpt-5.6-sol",
-      "totalModels": 56,
-      "trend": "flat",
+      "score": 91.2,
+      "scoreRaw": 0.912,
+      "modelName": "Kimi K3",
+      "modelId": "kimi-k3",
+      "totalModels": 57,
+      "trend": "up",
       "source": "llm-stats:browsecomp",
       "description": "Persistent web browsing and hard-to-find information retrieval."
     },
@@ -406,12 +406,12 @@
       "name": "DeepSearchQA",
       "category": "Search",
       "dimension": "toolSearchReliability",
-      "score": 93.1,
-      "scoreRaw": 0.931,
-      "modelName": "Claude Opus 4.8",
-      "modelId": "claude-opus-4-8",
-      "totalModels": 7,
-      "trend": "flat",
+      "score": 95,
+      "scoreRaw": 0.95,
+      "modelName": "Kimi K3",
+      "modelId": "kimi-k3",
+      "totalModels": 8,
+      "trend": "up",
       "source": "llm-stats:deepsearchqa",
       "description": "Multi-hop deep search and answer retrieval."
     }
@@ -566,6 +566,14 @@
       "bottleneckDimension": "agentsAutonomy",
       "updatedAt": "2026-07-16T05:14:47.391Z",
       "formula": "agi-readiness-v2"
+    },
+    {
+      "date": "Jul 17, 2026",
+      "agiReadiness": 76.1,
+      "agiDistance": 23.9,
+      "bottleneckDimension": "agentsAutonomy",
+      "updatedAt": "2026-07-17T05:17:41.505Z",
+      "formula": "agi-readiness-v2"
     }
   ],
   "legacyHistory": [