Google's Gemini 4 ships as four models; Carbon's rates set to double after Nov. 30
In brief
- Gemini 4 comes as four variants: Argon, Carbon, Hydrogen and Neon, per Crypto Briefing.
- All four share a 1-million-token context window but differ in reasoning depth and speed.
- Carbon and Argon move to $4 input and $20 output per million tokens after November 30.
- Hydrogen is priced at $1.50 input and $7.50 output per million tokens.
- Neon matched Carbon in coding evaluations at lower time and cost, Crypto Briefing says.
LeoDex News didn't see Google's own announcement. The lineup, positioning and pricing below come from Crypto Briefing's report.
Four models, one context window
Google has positioned each one for a different slice of work, from casual developer tinkering to enterprise-grade tasks.
Carbon is pitched as the most powerful option. It's tuned for complex problems that need deep reasoning, and in some consumer services it's restricted to the Ultra tier. Neon targets intricate tasks too, but with an emphasis on speed and cost-efficiency. Hydrogen is the budget option (built to run everyday jobs quickly at a lower price point).
Pricing flips on November 30
Gemini 4 is priced per million tokens, and output costs five times the input rate across all four models. Hydrogen sits at $1.50 input and $7.50 output, and its output rate is exactly half of Neon's.
Carbon's the odd one. Until the end of November it's actually cheaper per token than Neon, Crypto Briefing reported, but after the November 30 cutoff Carbon and Argon move to $4 per million input tokens and $20 per million output tokens, which doubles Carbon's rates and makes it the most expensive model in the lineup.
That's a real reshuffle for anyone running Carbon by default.
Reasoning effort is a second dial
Every Gemini 4 model supports adjustable reasoning effort: low, medium, high and, in some cases, xhigh. Carbon is optimized for medium effort and performs best of the group at low effort, while the others are optimized for high. At high effort, Neon and Hydrogen close the gap with Carbon and scale better, per the report.
Carbon reportedly scores about 65% on Terminal-Bench 4.0 for difficult tasks. Crypto Briefing also said Neon matched Carbon's results in coding evaluations in less time and at lower cost (both are the outlet's figures, not LeoDex News findings).
What does that mean for builders? Crypto Briefing's analysis is that some workloads that defaulted to Carbon may get rerouted to Neon or Hydrogen once the price change lands. Its broader point is that cost optimization now has two axes: which model you pick, and how much reasoning effort you apply.
Frequently asked questions
What are the four Gemini 4 models?
According to Crypto Briefing, Gemini 4 comes as four variants: Argon, Carbon, Hydrogen and Neon. All four share a 1-million-token context window. They differ in reasoning depth and execution speed, with Carbon pitched as the most powerful and Hydrogen built for everyday jobs at a lower price.
What changes for Gemini 4 Carbon pricing after November 30?
Crypto Briefing reported that Carbon is cheaper per token than Neon until the end of November. After the November 30 cutoff, Carbon and Argon move to $4 per million input tokens and $20 per million output tokens. That doubles Carbon's rates and makes it the most expensive model in the lineup.
How does reasoning effort affect Gemini 4 performance?
Gemini 4 models support low, medium, high and, in some cases, xhigh reasoning effort, per Crypto Briefing. Carbon is optimized for medium effort and performs best of the group at low effort. At high effort, Neon and Hydrogen close the gap with Carbon and scale better.


