{
  "id": 72063,
  "title": "DeepSeek's new AI model is by far the cheapest of well-known models to run, research firm says",
  "url": "https://urgent.news/2026/08/03/deepseeks-new-ai-model-is-by-far-the-cheapest-of-well-known-models-to-72063",
  "topic": "ai",
  "section": "AI",
  "published": "2026-08-03T05:43:40.000Z",
  "source": {
    "name": "Economic Times Tech",
    "slug": "economic-times-tech",
    "url": "https://economictimes.indiatimes.com/tech/artificial-intelligence/deepseeks-new-ai-model-is-by-far-the-cheapest-of-well-known-models-to-run-research-firm-says/articleshow/132820248.cms"
  },
  "original_language": "en",
  "account": "A Chinese startup, DeepSeek, has unveiled the V4-Flash version of its flagship AI model, which research indicates is the most cost-effective of the major players on the market, according to Artificial Analysis, a leading research firm. This news comes as DeepSeek prepares for a potential IPO in the near future. The V4-Flash model is a fraction of the cost to run compared to other leading models, including Anthropic's Claude Fable 5, which is 100 times more expensive. DeepSeek's R1 model gained global attention in early 2025, causing a significant drop in global technology stocks and raising questions about the substantial sums U.S. companies were investing in AI. The V4-Flash model charges $0.14 per million input tokens and $0.28 per million output tokens. For context, Kimi K3 from Moonshot AI costs 86 cents, OpenAI's GPT-5.6 Sol costs $1.86, and Anthropic's Claude Fable 5 costs $3.15. Artificial Analysis's evaluation of the V4-Flash model scored it 50 out of 100 on its Intelligence Index, placing it in the same league as Google's Gemini 3.6 Flash. However, it falls a point behind Meta's Muse Spark 1.1 and Z.AI's GLM-5.2. Despite this, DeepSeek continues to develop a more powerful version, the V4-Pro, whose release date remains undisclosed. Additionally, on the same day as the V4-Flash model's announcement, Alibaba unveiled its largest AI model, Qwen3.8-Max, which is yet to rival DeepSeek's offerings in terms of size and performance.",
  "summary": "A version of Chinese startup DeepSeek's flagship AI model is by far the least expensive to run on benchmark tests among well-known models globally and more than 100 times cheaper to run than Anthropic's Claude Fable 5, according to a research firm.",
  "key_points": [
    "DeepSeek's V4-Flash model is cheapest to run among major AI models.",
    "V4-Flash costs $0.14 per million input tokens, $0.28 per million output tokens.",
    "Artificial Analysis rates V4-Flash 50/100, comparable to Gemini 3.6 Flash."
  ],
  "editors_take": null,
  "illustration": null,
  "coverage": {
    "outlets": 3,
    "also_reported_by": [
      {
        "outlet": "Channel News Asia",
        "title": "DeepSeek's new AI model is by far the cheapest of well-known models to run, research firm says",
        "url": "https://urgent.news/2026/08/03/deepseeks-new-ai-model-is-by-far-the-cheapest-of-well-known-models-to",
        "published": "2026-08-03T05:41:43.000Z"
      },
      {
        "outlet": "South China Morning Post",
        "title": "China’s DeepSeek beefs up agentic AI with ‘harness’ tests as V4 model jolts Silicon Valley",
        "url": "https://urgent.news/2026/08/03/chinas-deepseek-beefs-up-agentic-ai-with-harness-tests-as-v4-model",
        "published": "2026-08-03T11:00:07.000Z"
      }
    ]
  },
  "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."
}