Argon, the Gemini 4 model, enabled the Chrome team to isolate core webpage handling from a ~6 million‑line legacy module within Chromium’s 40 million‑line codebase, allowing an architectural refactor across hundreds of files that compressed months of work into weeks and will let billions of users move live pages between tabs, side panels and floating windows without reloading. The 1‑million‑token output window lets the model execute 100 K+ token reasoning trajectories, achieving nearly 100× faster runtime and 90 % lower token consumption for 3D spatial reconstruction from video, turning raw footage into physics‑accurate scenes in seconds. In benchmarking, Argon ties for #1 on CWE‑Bench (68 % vulnerability remediation) and shows the strongest resilience against indirect prompt injections. It delivers a 14 % accuracy gain over Gemini 3.1 Pro, reaches 100 % success in long‑running evaluations, and completes complex multi‑step tasks 1.5× faster at a quarter of the cost per task in the Qualia autonomous researcher benchmark. Additional metrics include top‑ranked object detection, integration into Roboflow for high‑precision inspection, and 100 % successful completion of synthetic medical chart analyses with highest recall and ICD‑10 specificity. Safety mechanisms enforce refusal of harmful requests while monitoring chain‑of‑thought for misalignment.