{
  "id": 10484619,
  "title": "GPT-6 Sol vs. Claude Opus 5.5: Is cheaper important when results aren’t consistent?",
  "url": "https://urgent.news/2026/09/28/gpt-6-sol-vs-claude-opus-5-5-is-cheaper-important-when-results-arent",
  "topic": "ai",
  "section": "AI",
  "published": "2026-09-28T16:04:38.000Z",
  "source": {
    "name": "The New Stack",
    "slug": "the-new-stack",
    "url": "https://thenewstack.io/gpt-6-sol-vs-opus-5-5/"
  },
  "original_language": "en",
  "account": "On September 22, OpenAI unveiled GPT-6 Sol, touting that it outperforms Claude Opus 5 on business workflow tasks at a fraction of the cost per task. Anthropic released Opus 5.5 on the same day, so a test was conducted against the newer model. Pricing-wise, GPT-6 Sol costs $2 per million input tokens and $10 per million output tokens, which is half the price of GPT-5.6 Sol. Opus 5.5 costs $4 and $20.\n\nTo examine their performance, the author chose to test the models on consistency. Both models performed equally well on standard developer tasks, so the tests were rebuilt to measure consistency. Surprisingly, the models gave different results. While Sol was faster and cheaper, Opus 5.5 was more accurate across all tests. Overall, Sol cost about a sixth of what Opus 5.5 did, and was 8% cheaper on CI triage, the test closest to OpenAI's claim. Though Sol was faster and cheaper, Opus 5.5 was more accurate and consistent.",
  "summary": "OpenAI launched GPT-6 Sol on September 22, claiming it scores higher than Claude Opus 5 on business workflow tasks at The post GPT-6 Sol vs. Claude Opus 5.5: Is cheaper important when results aren’t consistent? 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."
}