AI Press Box
Language: English
EspañolPortuguêsDeutschFrançaisItalianoУкраїнська

Google doubles Gemini 4 Carbon and Argon API prices on 1 December 2026, to Claude Opus 5.5's $4 / $20

Published 17 November 2026. Prices as recorded on that date.

If your product runs on Gemini 4 Carbon or Argon, its token bill doubles in two weeks. As of 17 November 2026, Google's Gemini API pricing page lists Gemini 4 Carbon and Gemini 4 Argon at $2 per million input tokens and $10 per million output tokens "through November 30, 2026", and at $4 and $20 "starting December 1, 2026". That second price is exactly what Anthropic charges for Claude Opus 5.5.

The two Flash models follow a month later. Gemini 3.8 Flash and Gemini 3.6 Flash go from $0.75 / $3.75 to $1.50 / $7.50 on 1 January 2027.

What changes and when, as of 17 November 2026

All prices are from Google's Gemini API pricing page, paid tier, in US dollars per million tokens, input / output. Output includes thinking tokens.

ModelPrice nowNew priceFrom
Gemini 4 Carbon, Standard$2 / $10$4 / $201 December 2026
Gemini 4 Argon, Standard$2 / $10$4 / $201 December 2026
Gemini 4 Carbon and Argon, Batch and Flex$1 / $5$2 / $101 December 2026
Gemini 4 Carbon and Argon, Priority$3.60 / $18$7.20 / $361 December 2026
Gemini 3.8 Flash$0.75 / $3.75$1.50 / $7.501 January 2027
Gemini 3.6 Flash$0.75 / $3.75$1.50 / $7.501 January 2027

Gemini 4 Hydrogen, at $1.50 / $7.50, has no dated change.

Where the new prices sit

ModelInput / output per million tokens, as of 17 November 2026
Gemini 4 Carbon and Argon from 1 December 2026$4 / $20
Anthropic's Claude Opus 5.5$4 / $20
Anthropic's Claude Sonnet 5.5$2 / $10
Gemini 4 Hydrogen (no dated change)$1.50 / $7.50
Gemini 3.8 Flash from 1 January 2027$1.50 / $7.50
OpenAI's GPT-6 Astra$10 / $50

Until 30 November 2026, Gemini 4 Carbon and Argon cost the same as Claude Sonnet 5.5. From 1 December 2026 they cost the same as Claude Opus 5.5, and still less than half of OpenAI's GPT-6 Astra.

From 1 January 2027, Gemini 3.8 Flash costs the same as Gemini 4 Hydrogen, which Google describes as built for "everyday coding and agentic workflows". My reading is that a 3.8 model at a Gemini 4 model's price is a hint to move.

What it means for a team on Gemini 4 Carbon or Argon

Google prints both prices side by side with their dates, so to me this reads as an introductory price running out rather than a surprise hike. The bill doubles all the same.

In my view, a team that chose Gemini 4 over Claude Opus 5.5 on list price should run both on its own tasks before December, because from then on price no longer separates them. Total cost also depends on how many tokens a model spends per answer, thinking tokens included, and that differs by task.

Batch and Flex stay at half the Standard price after the change, $2 / $10. Work that can wait for an answer can absorb the increase that way.