
What happened
Google released Gemini 4 Argon today for developers, enterprises and subscribers, priced at $2 per million input tokens and $10 per million output tokens. It ranks #1 on the Vals Index and leads the Gray Swan Indirect Prompt Injection benchmark.
Why it matters
Argon is Google's most capable frontier model, aimed at high-stakes professional work like coding, finance, legal and cybersecurity, so its launch raises the bar for what businesses can expect from AI on complex, multi-step tasks.
WHO IT HITSEnterprise software teams and professional-services firms working on coding, legal, finance and cybersecurity tasks are the clearest beneficiaries. Early partners in Google's Fairwind Program, including Wiz, have already used Argon for security work.
Summaries like this, in your inbox every morning.
Gemini 4 Argon is the first step in the Gemini 4 family. It follows earlier Google models, and the company says Argon already outperforms other frontier models on benchmarks and has been used internally to increase agentic code submissions by more than 35% while reducing rollbacks by nearly a third.
The model launches alongside an outreach push. Google is offering 10,000 scientists the highest usage tier in Antigravity free for 6 months, and expanding its Fairwind Program to hundreds of additional partners. These moves suggest Google is positioning Argon not just as a product but as a platform for high-stakes professional work, from cybersecurity to enzyme design to robotics.
The pricing and safety details matter for adoption. With cached input tokens at 95% off and a 1M unified token limit, Argon appears designed for cost-sensitive enterprise workflows, while the safety stack addresses concerns about agent misuse and prompt injection that could otherwise slow corporate rollouts.
Pick your industry and the AI tools you use, and get news related to your work every day.
Free · 30 seconds with Google · unsubscribe anytimeWhat is AIToday? →
Ask AI anything about this article. The AI reads this article, earlier AIToday articles, and Wikipedia, and cites its sources. Q&As are published on this page for other readers too.
NEURA Robotics introduced Neuron AI, a multimodal foundation model combining 3D vision, touch, audio, robot st…

Google released Gemini 4 Argon globally for coding, knowledge work, enterprise workflows and cyber defense, wi…

Anthropic said its Japan staff grew from 3 to over 50 in a year, customer adoption rose 10-fold, and Claude us…

Anthropic announced Scion, a version-control system replacing Git, plus a code-review system, Antique

On October 12, 2026, Character.ai announced it began bankruptcy proceedings to sell certain assets to a Walt D…

NVIDIA may expand its stake in Reflection AI, an open-weight AI startup, a step linked to NVIDIA pushing furth…
