Key points
- Gemini 4 Argon is available globally via API, AI Studio, and Gemini apps.
- The model leads the Vals Index for performance in critical economic sectors.
- Introductory pricing starts at $2 per million input tokens and $10 per million output tokens.
The short version
- Google has released Gemini 4 Argon to developers, enterprises, and consumers worldwide.
- The model achieves number one on the Vals Index for finance, coding, legal, and tax tasks.
- Pricing includes significant discounts for cached input tokens, reducing costs by 95%.
The details
Google announced the global availability of Gemini 4 Argon today, making its latest frontier model accessible through multiple channels. Developers and enterprises can access the model via the API in AI Studio and the Gemini platform. Subscribers to Google AI Ultra can use it in Google Antigravity, while Ultra and Pro subscribers have access through the Gemini app and AI Mode in Search. This broad distribution targets professionals handling complex real-world tasks across coding, finance, legal, cybersecurity, and science domains.
According to Google, the model’s depth and precision in multi-step tasks allow it to outperform other frontier models on various benchmarks. It ranks first on the Vals Index, which measures performance in critical economic sectors including finance, coding, legal, and tax work. The launch includes an introductory price of $2 per million input tokens and $10 per million output tokens. Cached input tokens receive a 95% discount, offering a cost-effective option for large-scale processing.
The context
Google cited several internal and external use cases to demonstrate Argon’s capabilities. Software engineers at Google used the model to ship high-quality code faster, increasing agentic code submissions by more than 35% while reducing rollbacks by nearly a third. The Chrome team utilized Argon to separate core webpage handling from a six-million-line legacy folder within Chromium’s 40-million-line codebase, completing months of refactoring work in weeks. This change aims to eliminate browser crashes when moving webpages between tabs and windows.
In other engineering efforts, the team behind Google’s open-source Agent Development Kit used Argon to optimize their production Python codebase. The model cut in-memory session prep time by 40% and sped up cold imports by 21%. Researchers also applied Argon to the xTB quantum chemistry package, making it up to 2.3 times faster at predicting molecular structures. Additionally, Google reported a 2.85x end-to-end inference speedup for its Gemma models using Argon, achieving this result in days rather than months.
What to watch
- Monitor Google’s official channels for updates on pricing changes after the introductory period.
- Watch for new benchmarks or independent reviews that validate the Vals Index rankings.
- Observe how enterprise adoption impacts the availability of API resources and latency.
- Track any new security features or safeguards introduced for the model in future updates.
What it means for organisations
Organisations adopting this technology should assess how automated coding and analysis tools integrate with existing security and compliance workflows. Teams usually need to review vendor contracts for data handling clauses when using third-party AI services. Check whether your current infrastructure can support the increased throughput or if new resource allocations are required. Leadership should communicate clear guidelines on which tasks are suitable for AI assistance versus those requiring human oversight. Ensure that audit trails remain intact when AI agents modify code or process sensitive data.
What to do and how to stay safe: Google
- Review your current AI usage policies to ensure they cover generative models for coding and legal tasks.
- Verify that data sent to the API does not contain sensitive intellectual property or personal information.
- Monitor API costs closely, as usage can scale quickly with cached token discounts.
- Test the model’s outputs in a sandbox environment before integrating changes into production systems.
Step-by-step guide: Shadow AI Mistakes That Put Corporate Data at Risk
General security guidance from the TechSec Headlines newsroom. It is not confirmed advice from the organisations named in this story.
Frequently asked questions
How much does Gemini 4 Argon cost?
The introductory price is $2 per million input tokens and $10 per million output tokens, with cached inputs discounted by 95%.
Where can I access Gemini 4 Argon?
It is available globally via API, AI Studio, Google Antigravity, the Gemini app, and AI Mode in Search.
What benchmarks does Gemini 4 Argon lead?
It ranks first on the Vals Index for performance in finance, coding, legal, and tax work sectors.
