{
  "id": 5130652,
  "title": "Gemini 3.8 Flash and 3.8 Flash Cyber",
  "url": "https://urgent.news/2026/09/02/gemini-3-8-flash-and-3-8-flash-cyber-5130652",
  "topic": "ai",
  "section": "AI",
  "published": "2026-09-02T15:12:40.000Z",
  "source": {
    "name": "Hacker News Best",
    "slug": "hacker-news-best",
    "url": "https://blog.google/innovation-and-ai/models-and-research/gemini-models/3-8-flash-and-3-8-flash-cyber/"
  },
  "original_language": "en",
  "account": "Google has unveiled Gemini 3.8, their most advanced reasoning and coding model yet, following a rapid series of three releases in just six weeks. This new model retains the speed and affordability of its predecessor, 3.7 Flash, but brings significant improvements in coding and reasoning capabilities. Gemini 3.8 is powered by the same foundational intelligence but includes two variants to accommodate different deployment environments.\n\nThe enhanced reasoning and coding performance of Gemini 3.8 is largely attributed to rigorous training in cybersecurity. This model achieves substantial gains over previous versions, often matching the performance of more expensive frontier models, while being far more cost-effective. Gemini 3.8 Flash, in particular, demonstrates remarkable abilities in complex engineering problems, often outperforming larger models at a fraction of the cost. It also excels in quantitative and professional fields requiring advanced analysis and reporting, surpassing other frontier models in various benchmarks.\n\nWhat sets Gemini 3.8 apart is its ability to work harder when tackling complex tasks. It executes additional reasoning steps and iteratively calls tools, sometimes using more tokens to maximize performance. However, for applications where compute efficiency is paramount, developers can opt for lower effort levels to minimize token usage. For those in the cybersecurity field, Gemini 3.8 Flash Cyber offers a significant advantage. Available to trusted defenders through the Fairwind Program, this variant delivers frontier-level performance in autonomous vulnerability discovery, outperforming both 3.5 Flash Cyber and larger models. It achieves an impressive success rate of over 70% across a wide range of vulnerabilities in complex codebases spanning 20 programming languages. Notably, Gemini 3.8 Flash Cyber also prioritizes vulnerability fixing, offering a more permissive set of mitigations for cybersecurity while maintaining a cost-effective solution for trusted defenders. With an introductory price set to expire on December 31, 2026, Gemini 3.8 represents a major leap in AI capabilities, providing enhanced protection against malicious attacks while enabling beneficial use cases.",
  "summary": "https://deepmind.google/models/model-cards/gemini-3-8-flash/ Comments URL: https://news.ycombinator.com/item?id=49537553 Points: 388 # Comments: 232",
  "key_points": [],
  "editors_take": null,
  "illustration": null,
  "coverage": {
    "outlets": 2,
    "also_reported_by": [
      {
        "outlet": "Hacker News",
        "title": "Gemini 3.8 Flash and 3.8 Flash Cyber",
        "url": "https://urgent.news/2026/09/02/gemini-3-8-flash-and-3-8-flash-cyber",
        "published": "2026-09-02T15:44:06.000Z"
      }
    ]
  },
  "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."
}