Until November 30, 2026
$200.00
Introductory price
GOOGLE / API PRICING
Gemini 4 Argon Price Increase
Google launched Gemini 4 Argon at an introductory API price that ends on November 30, 2026. Enter your monthly usage to see what it costs before and after the standard price starts on December 1, 2026.
01 / YOUR MONTHLY USAGE
How much do you use Gemini 4 Argon?
Use millions of tokens: 2.5 means 2,500,000. Output includes any reasoning tokens your provider bills as output. The cached share is the part of your input billed at the cached-input rate; use 0 if you do not use context caching. Calculation runs in this tab; nothing is saved or sent.
02 / MONTHLY COST · USD
Gemini 4 Argon before and after
Your estimated monthly cost changes by +$200.00 (+100%).
From December 1, 2026
$400.00
Standard price
Monthly difference
+$200.00
+100%
50,000,000 uncached input + 0 cached input + 10,000,000 output tokens per month.
03 / RATES
What changes on December 1, 2026
| Rate | Until November 30, 2026 | From December 1, 2026 | Change |
|---|---|---|---|
| Input | $2.00 / 1M tokens | $4.00 / 1M tokens | +100% |
| Output | $10.00 / 1M tokens | $20.00 / 1M tokens | +100% |
| Cached input | $0.10 / 1M tokens | $0.20 / 1M tokens | +100% |
When does the price change?
Google's announcement of October 13, 2026 says Gemini 4 Argon launched at an introductory price and that the standard price of $4 input and $20 output per 1M tokens applies after November 30, 2026. The new price therefore applies from December 1, 2026.
Does the cached-input price change too?
Google's official Standard pricing table lists cached input at $0.10 / 1M tokens through November 30, 2026 and $0.20 / 1M tokens from December 1, 2026. These are published rates, not inferred prices. Gemini 4 Argon supports implicit caching only; enter the share actually billed as cached input, not a guaranteed cache-hit percentage.
How is the monthly cost calculated?
Uncached input tokens × the input rate, plus cached input tokens × the cached-input rate, plus output tokens × the output rate, all divided by 1,000,000. The cached share splits your input tokens between the first two. The same usage is priced at both rates, so the difference reflects the price change only.
What is not included?
- Batch or other service tiers, tool or grounding fees, subscription plans, committed-use or account discounts, and taxes.
- Any change in how many tokens your workload uses, or a different model choice.
- Prices or terms Google announces after October 26, 2026. Confirm current rates on the official pricing page before you budget.
NEXT / PATH