{
  "run": "2026-10-15T01-51-28-963Z",
  "models": [
    "alibaba/qwen3.8-max-prime",
    "anthropic/claude-fable-5.1",
    "anthropic/claude-haiku-5.5",
    "anthropic/claude-opus-5.5",
    "anthropic/claude-sonnet-5.5",
    "bggpt-gemma-3-27b-fp8",
    "deepseek/deepseek-v4-pro-0813",
    "deepseek/deepseek-v4.1-flash",
    "google/gemini-3.8-flash",
    "google/gemini-4-argon",
    "meta/muse-wonder-1.0",
    "mistral/mistral-large-4",
    "moonshotai/kimi-k3",
    "openai/gpt-5.6-luna",
    "openai/gpt-6-astra",
    "openai/gpt-6-luna",
    "openai/gpt-6.1-sol",
    "spacexai/grok-4.7",
    "zai/glm-5.3"
  ],
  "answers": 3610,
  "outcomes": {
    "answered": 3610
  },
  "answersUsd": 67.34,
  "judgingUsd": 16.9,
  "judges": [
    "openai/gpt-6.1-sol",
    "google/gemini-4-argon"
  ],
  "note": "Closed-book, one fixed Bulgarian system prompt (bench/run.mjs SYSTEM_PROMPT_BG); C4/C5 three runs each. Raw answers are kept locally (bench/runs is git-ignored); the public 30-item sample will be published with the results. Four Chinese frontier models were added on 15 Oct (founder); All models have a 32,000-token output limit; the 15 answers cut off at the original 8,000 were asked again.",
  "maxOutputTokens": 32000
}