{
  "id": 1508065,
  "title": "Codex vs. Claude Code at Liar's Dice: the Winning Bluff Was the Truth",
  "url": "https://urgent.news/2026/08/17/codex-vs-claude-code-at-liars-dice-the-winning-bluff-was-the-truth",
  "topic": "ai",
  "section": "AI",
  "published": "2026-08-17T15:14:49.000Z",
  "source": {
    "name": "Dev.to",
    "slug": "dev-to",
    "url": "https://dev.to/haoxiang_li_a709204042e6b/codex-vs-claude-code-at-liars-dice-the-winning-bluff-was-the-truth-203l"
  },
  "original_language": "en",
  "account": "In a high-stakes showdown, Claude Code emerged victorious over Codex CLI in three best-of-three matches of Liar's Dice, winning two matches each time. The highly engineered game was played using the same rules engine, with both AI agents utilizing distinct system prompts and tool loops. The entire series was run on an in-process rules engine with two stdio MCP servers, ensuring that neither side could see the other's dice and that the referee had no side channels favoring either AI. Claude Code's challenge calls hit 8 out of 11, while Codex's hit 4 out of 26. The score is not a general model ranking, but rather a result of the specific setup and execution of the game.",
  "summary": "One authoritative engine, two seat-locked MCP servers, three best-of-threes, and a 3-millisecond whodunit The matches are real: Codex CLI ( gpt-5.6-sol ) against Claude Code (Claude Opus 5), both playing through the same rules engine. Every number below was recomputed from the raw run.json and both session logs, and every game replays deterministically from its seed. Quotes from the agents are…",
  "key_points": [],
  "editors_take": null,
  "illustration": null,
  "coverage": {
    "outlets": 1,
    "also_reported_by": []
  },
  "ai_generated": true,
  "disclaimer": "Summaries, key points and the editor’s take are written by software from other outlets’ reporting and may contain errors — always check the linked original."
}