AI - Google released Gemini 4 Argon, is Google back? 🤣

Honestly, what a week it has been 🤣 We have every major player pretty much releasing something new. Google is the last one for now with their Gemini 4 Argon.

Initial response is that it is on par with GPT-6 Astra but a lot cheaper, so let us dive into what people are saying about it.

There are plenty of posts covering the event and the new announcements, so I will try to keep it relatively brief here 🤣

For my point of view, as someone who has private Gemini subscriptions because I have gmail and a lot of storage with Google, and now a bit of AI, I have been using their 3.6 models and those are utterly terrible, so I cannot image this one being any worse 🤣 I really hope this is a MASSIVE improvement as comparing their current models to anything from OpenAI, Anthropic or even Grok is night and day different. Not even in the same class at all.

Gemini 4 Argon

Gemini 4 Argon is already powering our internal workflows, with thousands of Googlers highlighting the model’s strengths in specialized coding tasks, conducting deeper research, and writing quality. It’s helping teams build faster and push the boundaries of engineering productivity and accelerating breakthroughs:

• Quantum algorithmic optimization: Argon is helping our quantum computing researchers optimize the spacetime resources (qubits × gates) of subroutines that bottleneck important applications. In one example, it beat the published baseline by 40% in a matter of minutes.

• Memory efficiency: A team of Argon agents analyzed fleet-wide profiling telemetry to autonomously identify and apply memory optimizations across Google’s data centers, freeing up over 300 TiB of memory once rolled out, with an estimated 500 TiB to 1 PiB in total savings.

• Large Scale Codebase Migrations and Optimizations: Argon agents are working on migrating C/C++ codebases to Rust across Google — scaling from tens of thousands of lines in core libraries like re2, libgav1 up to 800K+ lines for the Fuchsia Zircon kernel. Given the criticality of many of these systems, such large-scale rewrites are undergoing rigorous automated and manual auditing, emulation testing, and review before rolling out to production.

Some bold claims here, especially regarding quantum algorithmic optimisation and memory efficiency. But the large codebase migration is nothing to sneeze at either. If any of this is true, it is kind of cool 🤣

Official announcement

Gemini 4 Argon: our next era of frontier intelligence

Gemini 4 Argon delivers frontier performance in complex workflows across real-world software engineering, enterprise knowledge work like legal and finance, and cybersecurity defense.

Relevant X posts

Marketing

Pricing and performance

The super low hallucination rate is certainly impressive

1 million token OUTPUT

General thoughts

Bonus - Gemini App add skills

Automate tasks with skills

Automate your most repetitive tasks with tailored instructions through skills, which will be replacing Gems.

How was this not a thing before? 🤣

Useful links

• Introducing Gemini 4 Argon
• TechCrunch article on Gemini 4 Argon
• Automate tasks with skills