Particle.news

Google Rolls Out Gemini 4 Argon to Paying Developers and Subscribers

The company is banking on a unified one‑million token context window plus multimodal reasoning to speed multi‑step professional workflows.

Overview

  • Gemini 4 Argon, which Google began rolling out Tuesday, is now available worldwide to paid developers via the Gemini API and to Google AI Pro and AI Ultra subscribers in the Gemini app and Search AI Mode.
  • Google launched an introductory API price of $2 per million input tokens and $10 per million output tokens through November 30, after which those rates double to $4 and $20 per million respectively.
  • Argon uses a single one‑million token budget that counts both input and output together, meaning long prompts reduce room for model responses and changing how sustained multi‑step tasks are budgeted compared with rivals that separate context and output caps.
  • Access remains tiered and safety‑controlled: advanced agentic coding in Antigravity is limited to Ultra users, Google expanded its Fairwind program to give vetted cyber defenders an unguarded version for defensive work, and public releases include layered guardrails and runtime monitoring.
  • Google cites internal and partner gains — faster engineering refactors, higher agentic code submissions, and scientific and cybersecurity wins — but independent benchmarks give mixed results, leaving Argon’s comparative standing versus Anthropic and OpenAI contested even as Google pushes competitive pricing and enterprise trials.