Google releases Gemini 4 Argon frontier model
What happened
On September 30, 2026, The Decoder reported that Google released a new frontier model, Gemini 4 Argon, which matched OpenAI's GPT-6 Astra and Anthropic's Claude Fable 5.1 on several benchmarks but took no clear lead. The same day, AI HOT cited Artificial Analysis: in high-reasoning mode the model scored 53 on the Artificial Analysis Intelligence Index, level with GPT-6 Astra (max, 53) and one point ahead of GPT-6.1 Sol (max, 52). On October 1, Google DeepMind announced Gemini 4 Argon, saying it would first open through the Fairwind Program to trusted cyberdefense defenders, then expand to developers, enterprises and consumers. The same day The Verge reported Google's claim of frontier performance in software engineering, legal and financial enterprise knowledge work, and cyberdefense workflows, with initial access limited to a group of trusted security personnel. Google is taking part in the US government's voluntary pre-release model access process and widening access gradually, while strengthening safeguards against misuse, prompt injection and misalignment before broader release. On availability, earlier reports said the model opens first to trusted cyberdefense defenders and then expands to developers, enterprises and consumers; a later Ars Technica report said the model is currently used broadly only by the company's internal engineers and is not yet open to the public. The same day, AI HOT reported Artificial Analysis results: Gemini 4 Argon scored 53 on the Artificial Analysis Intelligence Index, matching GPT-6 Astra (max) and one point above GPT-6.1 Sol (max), at only 60% of the cost. Artificial Analysis then published an intelligence, performance and price analysis of Gemini 4 Argon (High) using Intelligence Index v4.3.2.
Written by AI from the coverage · updated 43 minutes ago
Developments
- Oct 1 00:00 · 3 reportsArtificial Analysis rates Gemini 4 Argon: Google back in the top three on intelligenceAI HOT · Models
- Sep 30 06:06 · 5 reportsGoogle DeepMind releases Gemini 4 Argon frontier modelGoogle DeepMind
Coverage
Follow the reports to see the story from different sides.
- Hacker News front pageGemini 4 Argon (High): Intelligence, Performance and Price Analysis
Artificial Analysis 发布 Gemini 4 Argon (High) 的智能、性能与价格分析，采用 Intelligence Index v4.3.2。
- The Verge · AIPickGoogle releases Gemini 4 Argon, initially limited to trusted cybersecurity staff
Google released its next frontier model, Gemini 4 Argon, saying it reaches frontier performance in complex workflows such as software engineering, legal and financial enterprise knowledge work, and cyber defense. The model opens first to a group of trusted cybersecurity personnel. Google is taking part in the US government's voluntary pre-deployment model access process and will widen access gradually, while strengthening safeguards against misuse, prompt injection and misalignment before a broader launch.
- AI HOT · ModelsArtificial Analysis 评测 Gemini 4 Argon，智能指数追平 GPT-6 Astra 且成本仅 60%
Artificial Analysis 评测 Google DeepMind 新模型 Gemini 4 Argon，在 Artificial Analysis Intelligence Index 得 53 分，追平 GPT-6 Astra (max)，高于 GPT-6.1 Sol (max) 1 分。该评测同时提到其成本仅为 60%。
- Ars Technica · AIGoogle announces Gemini 4 Argon AI model, but you can't use it yet
Google 发布 Gemini 4 Argon 模型，宣称在编码、知识工作和网络安全方面具备行业领先性能，但目前仅限公司内部工程师广泛使用，尚未对外开放。
- Hacker News front pagePickGoogle releases Gemini 4 Argon frontier model
Google released Gemini 4 Argon, its new frontier model, first through the Fairwind Program to trusted cybersecurity defenders, then gradually to developers, enterprises and consumers.
- Google DeepMindPickGoogle DeepMind releases Gemini 4 Argon frontier model
Google DeepMind released Gemini 4 Argon, its new frontier model. It goes first to trusted cyber-defense teams through the Fairwind Program, then widens to developers, enterprises and consumers. The release notes list an output cap, pricing and several benchmark scores.
- AI HOT · ModelsPickArtificial Analysis rates Gemini 4 Argon: Google back in the top three on intelligence
Artificial Analysis' evaluation shows Google DeepMind's Gemini 4 Argon scoring 53 on the Artificial Analysis Intelligence Index in high-reasoning mode, level with GPT-6 Astra (max, 53) and one point ahead of GPT-6.1 Sol (max, 52).
- The DecoderPickGoogle releases Gemini 4 Argon: benchmarks match OpenAI and Anthropic, no clear lead
Google released Gemini 4 Argon, its new frontier model. On several benchmarks it matches OpenAI's GPT-6 Astra and Anthropic's Claude Fable 5.1, but it does not take a clear lead.
Heat over time
Not enough continuous observations to draw a trend yet.