Google launches Gemini 4 Argon, a frontier model aimed at coding and professional work
3-line summary
- Google unveiled Gemini 4 Argon, its latest frontier model, on October 13 and released it globally the same day.
- The company says it is built for complex, multi-step work in coding, finance, law, cybersecurity and science.
- The introductory API price is $2 per million input tokens and $10 per million output tokens; in the Gemini app it is available to Google AI Pro and Ultra subscribers.
Google announced Gemini 4 Argon, a new frontier model, on its official blog on October 13. Developers and enterprises can use it through AI Studio and the Gemini API, and Google AI Pro and Ultra subscribers can use it in the Gemini app and AI Mode in Search. In Google Antigravity, it is available first to Google AI Ultra subscribers.
Google says Argon is built for deep work that takes many steps. The company claims it ranks first on the Vals Index, which measures model performance in sectors such as finance, coding, legal and tax work. Internally, Google says using Argon in Antigravity increased agent-submitted code by more than 35% and cut code rollbacks by nearly a third.
Google also shared examples. The Chrome team used Argon to separate a legacy folder of about 6 million lines from Chromium's 40-million-line codebase, finishing months of work in weeks, and Argon sped up inference for the open Gemma 4 E2B model by 2.85x in a few days. Through the Gemini API, input and output together can reach 1 million tokens.
Most of the performance figures and examples come from Google and its early-access customers, and the current price is an introductory one. Google also says Argon is better than earlier Gemini models at refusing dangerous requests and at resisting prompt injection — malicious instructions hidden in webpages or emails.
Why it matters
By pricing a model aimed at the hardest coding and professional tasks at an introductory $2 per million input tokens, Google looks set to sharpen price-performance competition among frontier models. Its emphasis on agents that keep working through long tasks on their own also stands out.
Who it's for
- Developers
- It's available now on the Gemini API paid tier. While the introductory price lasts, run your current workloads through it and compare cost and results.
- Office workers
- If you subscribe to Google AI Pro or Ultra, you can try it today in the Gemini app and AI Mode in Search. Start with time-consuming work such as analyzing long documents.
