Google releases Gemini 4 Argon, tops benchmarks in coding and cybersecurity
Google announced Gemini 4 Argon on Wednesday, calling it its most advanced AI model yet. The company claims it sets new records in real-world software engineering, ties for first in cybersecurity, and leads benchmarks for professional tasks like finance and legal work. Internally, Argon has already reduced memory usage at Google’s data centers by hundreds of terabytes without extra hardware and…
Key points
- Google claims Gemini 4 Argon tops benchmarks in coding, cybersecurity, and professional tasks like finance and legal work
- Internally, Argon freed hundreds of terabytes of memory at Google’s data centers without new hardware
- Public rollout delayed for safety reviews with cybersecurity partners and the U.S. government
The model will roll out in phases, starting with cybersecurity partners and U.S. government safety evaluations. Google’s Tulsee Doshi said this phased approach ensures safeguards against misuse and prompt injection before public launch. The release follows a voluntary AI safety accord signed by CEO Sundar Pichai with President Trump the day prior.
Model pages: Gemini 4 Argon → · GPT-6 Astra → · Grok 4.7 →
The story so far
2 episodes →- Google releases Gemini 4 Argon, tops benchmarks in coding and cybersecuritythis story
Google rolls out Gemini 4 Argon, its most advanced AI model
CNBC Technology · 30 September 2026
Loading the full article…
This text was published by CNBC Technology and written by MacKenzie Sigalos,Samantha Subin. It is reproduced here with attribution so you can read it in full; the rights remain with the publisher. Read it at the source ↗
Coverage and discussion
4sources- Google announces Gemini 4 and says it’s so capable that only ‘trusted cyber defenders’ can have it right nowPress · The Verge AI ·
- Sources: some Google employees say Gemini 4 performs well on benchmarks but struggles with some real-world coding tasks; Google disputes that characterizationPress · bloomberg.com ·
- Google rolls out Gemini 4 Argon to a small group of cybersecurity partners and says it outperforms GPT-6 Astra on certain coding and knowledge work benchmarksPress · axios.com ·
The headline, key points and digest above were generated by Digest AI's editorial model from the linked sources. Automated summaries can contain errors: the sources are the record. Spotted a mistake? Tell us. Published by Martin K., who runs Digest AI and handles corrections.
More in Generative AI & Models
All →- OpenAI releases GPT-6.1 Sol, nearly matching Astra at one-fifth the price · 24 src
- Google announces Gemini 4 Argon frontier model · 2 src
- OpenAI scraps GPT-6.1 Astra release over safety concerns · 61 src
- InclusionAI releases Ling 3.1 Flash with 2-week free trial · 1 src
- Anthropic launches Claude Sonnet 5.5, claiming 30% faster output and up to 30% lower cost per task than Sonnet 5 · 25 src
Comments
via GitHub Discussions