Gemini 4 Argon (Google AI Studio)

Google made Gemini 4 Argon generally available on Google Cloud on October 13, 2026, framing it as a frontier model trained for depth and precision so it can keep context across long, multi-step tasks reinforced by strong domain expertise. Google Cloud targets enterprise workloads including complex software builds, detailed financial and legal work, and cybersecurity defenses such as CodeMender, where teams need sustained focus from start to finish. Google says Argon turns huge, unstructured inputs mixing text, tables, charts, and images into one grounded, structured output, can analyze data over time, and is trained to respect system boundaries, report negative findings honestly, and pause for confirmation before irreversible steps. The post positions the model for handing off the deep, multi-step work that previously took hours or days, returning finished, usable work rather than draft answers.

Google globally launched Gemini 4 Argon on October 13, 2026, the first Gemini 4 generation model, available via the Gemini API in AI Studio and the Gemini platform for developers and enterprises, in Antigravity for AI Ultra subscribers, and in the Gemini app and Search AI Mode for AI Ultra and Pro subscribers. Google positions it as a frontier model excelling at coding, finance, legal, cybersecurity, and science, ranking first on the Vals Index across those domains, with depth and precision over long, multi-step workflows. Introductory pricing is $2 per million input tokens and $10 per million output tokens, with cached input priced 95% off, running through November 30, 2026 before stepping up to $4 and $20. Google reports internal usage where Argon in Antigravity raised agentic code submissions by more than 35% and cut rollbacks by nearly a third, including a multi-month Chromium refactor condensed into weeks. Early access also covered finding and patching critical cybersecurity vulnerabilities and progress in quantum, fusion, and life sciences research.

Google announced Gemini 4 Argon on September 30, 2026 as its next-era frontier intelligence model built for sustained deep reasoning across long-horizon, real-world software engineering, enterprise knowledge work like legal and finance, and cybersecurity defense. The launch uses a phased rollout, beginning with trusted cyber defenders in Google's Fairwind Program under the U.S. government's voluntary pre-release access process, before broader release. Argon features an industry-leading 1 million token output limit for deep multi-step problem solving, versus 64,000 on earlier Gemini models, and launches at $2 per million input tokens and $10 per million output tokens with cached input tokens at 95% off. Google says thousands of internal users rely on Argon for specialized coding, deeper research, and writing quality, and that the model can autonomously find, validate, and patch critical software vulnerabilities across large codebases.