Google's Gemini API price page lists two of its newest models at launch prices that double on fixed dates: Gemini 4 Argon on 1 December 2026, and Gemini 3.8 Flash, with the older 3.6 Flash, on 1 January 2027. For an ecommerce brand running AI on product content or customer support, that doubles the model line of the bill unless the work moves to Batch pricing or to a cheaper model. This guide uses Google's own price page, dated 22 October 2026 and read on 26 October 2026, and is from Brand Media Works, an AI consulting and marketing data engineering firm for ecommerce brands.
What are Gemini API prices before and after the change?
| Model | Tier | Price today | Price after the change | Change date |
|---|---|---|---|---|
| Gemini 4 Argon | Standard | $2.00 / $10.00 | $4.00 / $20.00 | 1 December 2026 |
| Gemini 4 Argon | Batch and Flex | $1.00 / $5.00 | $2.00 / $10.00 | 1 December 2026 |
| Gemini 4 Argon | Priority | $3.60 / $18.00 | $7.20 / $36.00 | 1 December 2026 |
| Gemini 3.8 Flash | Standard | $0.75 / $3.75 | $1.50 / $7.50 | 1 January 2027 |
| Gemini 3.8 Flash | Batch and Flex | $0.375 / $1.875 | $0.75 / $3.75 | 1 January 2027 |
| Gemini 3.6 Flash | Standard | $0.75 / $3.75 | $1.50 / $7.50 | 1 January 2027 |
| Gemini 3.5 Flash-Lite | Standard | $0.30 / $2.50 | No change listed | |
| Gemini 3.1 Flash-Lite | Standard | $0.25 / $1.50 | No change listed | |
| Gemini 3.1 Pro Preview | Standard, prompts up to 200k tokens | $2.00 / $12.00 | No change listed |
Context caching doubles on the same dates. Gemini 4 Argon is not available on the free tier. In April 2026, an Internet Archive copy of the same page listed Gemini 3 Flash Preview at $0.50 / $3.00 as Google's "most intelligent model built for speed"; it is still listed at that price, now described as Google's legacy Flash model.
What does the change cost an ecommerce brand?
Take a catalog job: rewriting and enriching descriptions for 10,000 products, with about 2,000 tokens of product data and brand guidelines in and 400 tokens out per product. That is 20 million input and 4 million output tokens each time it runs.
| Model and tier | Today | After the change |
|---|---|---|
| Gemini 4 Argon, Standard | $80 | $160 from 1 December |
| Gemini 4 Argon, Batch | $40 | $80 from 1 December |
| Gemini 3.8 Flash, Standard | $30 | $60 from 1 January |
| Gemini 3.8 Flash, Batch | $15 | $30 from 1 January |
| Gemini 3.5 Flash-Lite, Standard | $16 | $16 |
| Gemini 3.1 Flash-Lite, Standard | $11 | $11 |
The amounts are small for one run. They grow with work that runs every day: a support team drafting replies to 20,000 tickets a month, at 60 million input and 10 million output tokens, would pay $220 a month on Gemini 4 Argon today and $440 from 1 December (our arithmetic). AI model API prices in 2026 prices the same support workload across DeepSeek, Qwen, Zhipu and OpenAI.
How does Gemini compare with OpenAI and Anthropic after the change?
| Model | Price |
|---|---|
| Gemini 4 Argon, from 1 December 2026 | $4.00 / $20.00 |
| Anthropic Claude Opus 5.5 | $4 / $20 |
| OpenAI gpt-6.1-sol | $2.00 / $10.00 |
| Anthropic Claude Sonnet 5.5 | $2 / $10 |
| Gemini 3.8 Flash, from 1 January 2027 | $1.50 / $7.50 |
| OpenAI gpt-6-luna | $0.10 / $0.50 |
| Anthropic Claude Haiku 5.5, prompts up to 100,000 tokens | $0.10 / $0.50 |
Price per token is only part of the comparison. Models differ in quality on a given task and in how many tokens they produce for the same text; Anthropic notes that its models from Claude 4.7 onward use a tokenizer that produces about 30% more tokens. The fair test is the cost of a real workload at the quality the brand needs.
What should a brand using Gemini do before 1 December?
- Find out which model each workflow calls. A workflow pinned to Gemini 4 Argon or 3.8 Flash doubles in cost on the change date without anyone changing it.
- Move work that can wait to Batch. Catalog rewrites, product tagging and review summaries rarely need an answer in seconds. At Batch prices after the rise, they cost what they cost on Standard today.
- Test a Flash-Lite model on simple tasks. Classification, extraction and short rewrites may run well on Gemini 3.5 or 3.1 Flash-Lite, which show no scheduled change.
- Budget for search grounding. Google gives 5,000 free search requests a month across Gemini 3 and newer models, then charges $14 per 1,000, and one request can trigger several searches.
- Keep the model swappable. A workflow built so the model can be changed in one place can move to another provider when prices shift. Private LLM vs OpenAI API covers when running an open-weight model in the brand's own account is cheaper and more predictable.
Private LLMs on open-weight models describes how Brand Media Works runs models inside a brand's own AWS account at a cost that does not move with a vendor's price list, and what a private LLM costs to run sets out the numbers to compare.
Frequently asked questions
When does Gemini API pricing go up?
Gemini 4 Argon prices double on 1 December 2026. Gemini 3.8 Flash and Gemini 3.6 Flash prices double on 1 January 2027. Google lists both dates on its Gemini API price page.
How much does Gemini 3.8 Flash cost?
On the Standard tier, $0.75 per million input tokens and $3.75 per million output tokens through 31 December 2026, then $1.50 and $7.50 from 1 January 2027. Batch prices are half of those.
Is the Gemini API free?
There is a free tier for some models, but Gemini 4 Argon is not on it, and Google says free-tier content is used to improve its products. Paid-tier content is not, which matters for any prompt that carries customer data.
What is the cheapest Gemini model?
Of the models on Google's price page, Gemini 2.5 Flash-Lite has the lowest text price, at $0.10 per million input and $0.40 per million output tokens. Among Gemini 3 models, Gemini 3.1 Flash-Lite is lowest at $0.25 and $1.50.