{
  "id": 9175231,
  "title": "OpenAI launches GPT-6 Sol and Luna, saying Sol makes about half as many mistakes as GPT-5.6 Sol and Luna matches GPT-5.6 Sol's performance at ~1% of the cost (David Gewirtz/ZDNET)",
  "url": "https://urgent.news/2026/09/22/openai-launches-gpt-6-sol-and-luna-saying-sol-makes-about-half-as",
  "topic": "ai",
  "section": "AI",
  "published": "2026-09-22T18:15:08.000Z",
  "source": {
    "name": "Techmeme",
    "slug": "techmeme",
    "url": "https://www.zdnet.com/innovation/openai-gpt-6-sol-luna-release/"
  },
  "original_language": "en",
  "account": "OpenAI has unveiled its latest mainstream AI models, GPT-6 Sol and GPT-6 Luna, boasting significant improvements in reliability and cost-effectiveness. According to the company, GPT-6 Sol makes about half as many mistakes as its predecessor, GPT-5.6 Sol. In contrast, GPT-6 Luna matches the performance of GPT-5.6 Sol at roughly one-tenth the cost. The launch of these models comes amidst a fierce competition between OpenAI and Anthropic in the AI market.\n\nIn simple terms, the new GPT-6 models are set to revolutionize the AI landscape by offering better accuracy and affordability. While both OpenAI and Anthropic provide flagship, workhorse, and budget-friendly model options, these new releases position OpenAI as a formidable competitor. The rate of improvement in these models is astonishing, with OpenAI managing to double the factual accuracy of workhorse models in just three months. This achievement is comparable to elevating the performance of top doctors to that of super-experts and upgrading low-tier residents to match the skills of top doctors.\n\nBenchmarking these new models against their competitors reveals significant gains. For instance, GPT-6 Luna outperforms Anthropic’s Claude Opus 5 by 6.3% at just 9% of the cost per task. Similarly, GPT-6 Sol surpasses Anthropic’s Claude Fable 5.1 xhigh by a slim margin of 1.1 percentage points while being 80% more cost-effective. In complex professional workflows and coding tasks, GPT-6 Sol demonstrates impressive results that rival their counterparts, albeit at a significantly lower cost.\n\nThe standout performance of GPT-6 Luna is particularly noteworthy. It achieves comparable results to Anthropic’s Opus 5 and Fable 5 models at medium effort while being 93% less expensive per task. OpenAI's announcement highlights the vast potential of these new models, emphasizing the cost advantages that could transform the AI landscape. However, the company remains tight-lipped about the specific benchmarking details, leaving room for speculation and further analysis.",
  "summary": "GPT-6 Sol and Luna arrive less than three months after GPT-5.6, with OpenAI claiming major accuracy gains, dramatic cost reductions …",
  "key_points": [],
  "editors_take": null,
  "illustration": null,
  "coverage": {
    "outlets": 4,
    "also_reported_by": [
      {
        "outlet": "TechCrunch",
        "title": "OpenAI launches GPT-6 Sol and Luna, boasting lower cost and fewer mistakes",
        "url": "https://urgent.news/2026/09/22/openai-launches-gpt-6-sol-and-luna-boasting-lower-cost-and-fewer",
        "published": "2026-09-22T18:00:00.000Z"
      },
      {
        "outlet": "The New Stack",
        "title": "OpenAI releases GPT-6 Sol and Luna — and cuts token prices in half",
        "url": "https://urgent.news/2026/09/22/openai-releases-gpt-6-sol-and-luna-and-cuts-token-prices-in-half",
        "published": "2026-09-22T18:00:02.000Z"
      },
      {
        "outlet": "Dev.to",
        "title": "OpenAI Cuts GPT-5.6 Luna and Terra API Prices, Expanding Room for Automation",
        "url": "https://urgent.news/2026/09/22/openai-cuts-gpt-5-6-luna-and-terra-api-prices-expanding-room-for",
        "published": "2026-09-22T20:30:31.000Z"
      }
    ]
  },
  "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."
}