{
  "id": 140212,
  "title": "OpenAI’s Astra just proved 10 long-standing math and science theorems. The tokens cost $2,000.",
  "url": "https://urgent.news/2026/08/04/openais-astra-just-proved-10-long-standing-math-and-science-theorems",
  "topic": "ai",
  "section": "AI",
  "published": "2026-08-04T17:57:06.000Z",
  "source": {
    "name": "The New Stack",
    "slug": "the-new-stack",
    "url": "https://thenewstack.io/openai-astra-math-cost/"
  },
  "original_language": "en",
  "account": "OpenAI unveiled this week that an internal version of its Astra model, a next-generation AI system, has independently generated machine-verified proofs for 10 longstanding math and computer science problems for roughly $2,000 in GPT-5.6 Sol API tokens. This pricing structure sets a new benchmark for developers to consider when allocating resources for AI-powered research. While the work is not yet complete and requires human review, the announcement offers a clearer picture of the potential costs associated with accessing frontier AI reasoning. OpenAI has not disclosed the details of the model or its pricing, but the $2,000 estimate is based on standard token costs for GPT-5.6 Sol. The company emphasizes that the research process generated 10 new results on long-standing open problems, providing developers with a new opportunity to evaluate the value of frontier AI capabilities. Researchers acknowledge that the model's performance on more complex problems like the Millennium Prize Problems remains uncertain, but the demonstration highlights the potential for democratizing access to cutting-edge AI technology.",
  "summary": "OpenAI shared a big research update this week, announcing that an internal version of Astra, its next frontier model, created The post OpenAI’s Astra just proved 10 long-standing math and science theorems. The tokens cost $2,000. appeared first on The New Stack .",
  "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."
}