Google has unveiled its next flagship AI model: Gemini 4 Argon, launched on October 14, 2026, and aimed squarely at professional workloads in coding, finance, law, and science. The release is picking up viral traction on Instagram and Threads, where reactions to the new flagship are spreading fast.

Pricing is set at $2 per million input tokens and $10 per million output tokens — but Google says rates will double after November 30, so teams planning heavy API usage may want to budget accordingly. Pricing changes often in the AI space, so it's worth confirming current rates on Google's official pricing page before building around them.

The model is available now to Pro and Ultra subscribers as well as through Google's API, targeting both developers and businesses. The dual-track rollout — consumer subscription plus enterprise API — mirrors how Google has shipped recent Gemini generations, widening the funnel from individual power users to production deployments.

Gemini 4 Argon lands in the middle of a crowded frontier-model season, with every major lab racing to own the "professional workloads" narrative. The domain targeting is the notable choice here: rather than one generalist model for everything, Google is positioning Argon explicitly for work where accuracy has business consequences.

What remains unproven at launch: independent benchmark results for coding, legal, and scientific reasoning beyond Google's own claims. Until third-party evaluations land, treat the flagship label as positioning rather than proof. Still, the combination of domain targeting and introductory pricing makes Argon one of the quarter's most closely watched releases.

Sources