Overview
- Google rolled out Gemini 4 Argon on Tuesday, making the model available worldwide to developers via the paid Gemini API and to enterprise and paid Google AI Pro and Ultra subscribers in the Gemini app, Search AI Mode and Antigravity.
- Argon uses a single one million‑token budget that counts both input and output, so very large prompts will reduce how much the model can write back and users must trade prompt length against response length.
- Google set an introductory API price of $2 per million input tokens and $10 per million output tokens through November 30, with rates scheduled to double after that date.
- The company expanded its Fairwind program to give vetted cybersecurity defenders an unguarded Argon for vulnerability hunting while retaining guardrails, real‑time agent monitoring and prompt‑injection defenses for public users.
- Google cites top benchmark placements and internal gains such as faster agentic code submissions, but independent third‑party evaluations remain limited and some employees reported coding weaknesses that Google disputes, which leaves questions about real‑world performance and total user cost.