Google Launches Gemini 4 Argon With Aggressive Pricing for Businesses

Google Launches Gemini 4 Argon — With Aggressive Pricing

After months trailing its rivals, Google rolled out its new flagship model, Gemini 4 Argon, broadly today (Oct 13), claiming it competes with the industry's best — including GPT-6 Astra and Anthropic's Claude models — on professional tasks like coding, finance, law, and cybersecurity.

What really stands out is the pricing: Google is launching Argon at an introductory rate of $2 per million input tokens and $10 per million output tokens — about a fifth the cost of OpenAI's GPT-6 Astra. The price doubles only at the start of December. Google reports that more than 100,000 of its own employees already use the model internally, with a 35% increase in automated code submissions and a 30% drop in rollbacks.

The model is now available to Pro and Ultra subscribers in the Gemini app, AI Mode in Search, and the Antigravity agentic development platform. The Gemini app is also gaining expanded agentic capabilities — including background tasks that keep running while users are away from their computer, and connections to more than 40 external apps.

Why it matters for small and medium businesses: significantly lower per-token pricing lowers the barrier to experimenting with frontier-grade AI capabilities, without committing to premium pricing from competitors.

Source: The Verge, VentureBeat

Shopify Opens Its Catalog to Any AI Shopping Agent

Shopify announced that any AI agent — not just Meta's Muse, with which it partnered in September — can now access over a billion listings in its catalog and complete purchases on a customer's behalf, through a connection to the Shop app and one-click Shop Pay checkout. Customers can set budget limits and preferences, and still must approve each individual purchase.

Why it matters: for small and medium Shopify store owners, this is a meaningful expansion of exposure to the growing ecosystem of AI shopping assistants — not locked into a single partner integration.

Source: The Verge

Amex: A New Safeguard Against AI Agent Purchase "Mistakes"

American Express launched its "Intent Intelligence Framework" — a system that captures and authenticates a customer's exact purchase intent (including merchant name, maximum budget, and preferences) before an AI agent completes a transaction, aiming to prevent cases where an agent "misfires" and buys the wrong product.

Why it matters: as more businesses and consumers let AI agents make purchases autonomously, this kind of control mechanism becomes critical for preventing financial harm and building trust.

Source: TechCrunch

---

Just as American Express is building a verification and control layer before an AI agent executes a financial action, the same principle applies to office automation — the right approach is to let AI agents actually carry out tasks, but with clear oversight and human approval where it matters. That's exactly the philosophy guiding RealOffice.AI's development of automation agents for small and medium offices.