AI news · October 14, 2026
Gemini 4 Argon opens globally at $2 per million input tokens
Google opened Gemini 4 Argon globally at $2/M input and $10/M output, with cached input 95% cheaper; builders can run one fixed task in AI Studio before comparing it with their current model.
- introductory input price
- $2/M
- introductory output price
- $10/M
- cached input discount
- 95%
Google says Gemini 4 Argon is now globally available to developers and enterprises through the API in AI Studio and the Gemini platform. Paid Google AI users can also use it in Gemini, AI Mode in Search and Google Antigravity, depending on their plan. Google positions Argon for long, difficult work in coding, finance, legal work, cybersecurity and science, and says it ranks first on the Vals Index. Those performance claims come from Google and need a task-level check.
The launch price is the practical change: $2 per million input tokens and $10 per million output tokens, with cached input tokens discounted by 95%. The introductory rate is scheduled to rise after November. A maker can now compare one fixed research or coding task across Argon and an existing model, recording correct completion, tool errors, time and total spend. The cheaper token rate is useful only if the model finishes the job with similar review effort.
What you can do with it
Open AI Studio and run a small, repeatable task with the same prompt, tools and input used for your current model. Record completion rate, retries, latency, input tokens, output tokens and human corrections. Keep the old route active until Argon proves a lower total cost on your own work.
Our take
Argon is a meaningful follow-up because access and price are now concrete rather than promised. The headline rate is attractive, but a model can spend more tokens or need more review than a rival. The best near-term use is a controlled side-by-side run on work where quality and time are easy to score.
Source: Google — Made With Models writes the brief; the reporting is theirs.