Before Gemini’s promo rates rise: Argon goes up at the end of November, Flash at year end

The scene

You use, or plan to use, the Gemini API for reasoning or agent work, and you want to know the current rates, whether a free tier still exists, and whether older model names will be rerouted — before the dates Google has already published.

Tools involved

What they did

1. Split flagship reasoning from everyday agent work. The official page says Gemini 4 Argon (gemini-4-argon) is not available on the free tier, and that every token it processes is billed, including thinking tokens and the internal tool instructions and function declarations used with grounding.
2. If you need Argon: the paid list price through 30 Nov 2026 is $2.00 input and $10.00 output (including thinking) per million tokens, with context caching at $0.10. From 1 Dec 2026 those figures become $4.00, $20.00, and $0.20. The page says only implicit caching is supported, not explicit caching. Comparisons or small trials at the current rate need to finish before 30 November ends.
3. For everyday agents and longer coding jobs, look at Gemini 3.8 Flash. The free tier lists input and output as free of charge. Paid rates through 31 Dec 2026 are $0.75 input, $3.75 output, and $0.075 for cache, plus $0.50 per million tokens per hour for cache storage. From 1 Jan 2027 the paid rates become $1.50 / $7.50, cache $0.15, and storage $1.00. The free tier is marked as used to improve Google’s products; the paid tier is not.
4. For high-volume, well-specified work, Gemini 3.5 Flash-Lite lists free input and output on the free tier, and paid rates of $0.30 input / $2.50 output, with cache at $0.03 and storage at $1.00 per million tokens per hour.
5. If you still call older names: the page says Gemini 2.5 Pro is deprecated, and that from 1 Dec 2026 requests to that name are routed to Gemini 3.8 Flash and billed at 3.8 Flash rates. Gemini 2.5 Flash is rerouted the same day to Gemini 3.5 Flash-Lite and billed at those rates.
6. The paid tier lists a Batch API at 50% lower cost. For Grounding with Google Search on Gemini 3 and newer models, the page lists 5,000 free requests per month (shared across those models), then $14 per 1,000 requests, with a default developer-API cap of 1,500 requests per day. Use the official page for any other free or paid limits; this site does not invent counts that are not written there.

Copyable prompt

I will paste official Gemini prices and the dates next to them. Tell me which model I should use for [kind of work], what I will pay at today’s rates, and which of those rates change on which date. If a figure or date is missing from the text I pasted, say so and do not invent it. Also warn me if an older model name I mention is scheduled to be rerouted.

Rewritten here for you to try; not a verbatim excerpt from elsewhere. What it asks：Paste official Gemini prices and dates, and ask which model fits the job, what it costs at today’s rates, which rates change when, and whether an older model name will be rerouted.

Things to watch

These figures come from the Google AI Studio / Gemini Developer API pricing page checked on 10 Nov 2026. They are USD list prices per million tokens, before tax. Thinking tokens are billed as output. The free tier may use content to improve Google’s products, and some models are not available there. The page notes that Gemini Enterprise Agent Platform prices may differ. If the official page changes, use the new page.