Gemini 4 Argon’s API price is a limited-time deal. Google lists $2 per million input tokens and $10 per million output tokens (USD) through November 30, 2026, and $4 and $20 starting December 1. If you plan to build on it, budget at the higher price, not the intro one. This is an analysis of Google’s documentation, not a hands-on test.

What Google has published
Google’s Gemini API model page says the model is “Available through Nov 30, 2026 at an introductory price of $2/1M input tokens and $10/1M output tokens.” The pricing page shows the same split for the standard tier: $4.00 for input starting December 1, 2026, with output rising to $20.00. Context caching moves from $0.10 to $0.20 per million tokens on the same date.
Google’s October 13 developer post describes the model as built for long software and enterprise workflows, and says to start on the paid tier of the Gemini API in Google AI Studio and Android Studio.

The numbers side by side
| Item (USD, per 1M tokens) | Through Nov 30, 2026 | From Dec 1, 2026 |
|---|---|---|
| Input | $2.00 | $4.00 |
| Output (including thinking) | $10.00 | $20.00 |
| Context caching | $0.10 | $0.20 |
A quick example, using my own arithmetic: a job that uses 10 million input tokens and 2 million output tokens costs $40 at the intro rate and $80 at the standard rate. Real bills also depend on retries and thinking tokens, which count as output.
Who it matters for
- Teams evaluating now: run your trial before December 1 if you want cheap data, but project costs at double.
- Anyone planning production use: the intro price is not a stable baseline. Compare it with other models using total spend per accepted job, as in our look at GPT-6.1 Terra pricing and Kimi K3.1 Pro.
- Hobbyists: Google’s model page says “Gemini 4 Argon is not available on the Free tier.”
Limits and what is still unknown
Google’s pages describe one window of 1,048,576 tokens shared by input, thinking and output, so long prompts leave less room for answers. Prices outside the US dollar listing were not checked. I did not test quality, speed or the real cost per task, and I found no independent benchmark that I trust enough to cite. Google has not said whether the date could move, so recheck the pricing page before budgeting. For what the model’s showcase demos do and do not prove, see our piece on Gemini 4 Argon’s 3D demos.
Bottom line
Treat $4 and $20 as the real price. Use the intro window to test on your own tasks, record retries, and decide before December 1 whether the model earns that rate.

