Google's Gemini 4 Neon undercuts Carbon on price
Google released Gemini 4 Neon on Wednesday at $3 per million input tokens and $15 per million output, with a 1 million token context window. Once Carbon's price doubles on Dec. 1, Neon is the Gemini most builders should test first.
- BY
- TopFive Desk
- FILED
- READ
- 3 min
Google's newest Gemini sits in the middle of its lineup. That is where a lot of real work runs.
Gemini 4 Neon shipped on Wednesday, and the price tag is the part to read first.
Neon costs less than Carbon and Google says it codes better
Google released Gemini 4 Neon on Nov. 18, a model it pitches at software engineering, knowledge work and science. In Google's lineup it sits between Gemini 4 Carbon, built for the hardest long-horizon work, and Gemini 4 Hydrogen, built for fast everyday tasks.
The Gemini API price list puts Neon at $3 per million input tokens and $15 per million output tokens, with cached input at $0.15. The context window is 1 million tokens.
$3 in, $15 out.
Carbon is cheaper today only because of its introductory price. The same page lists Carbon at $2 and $10 through Nov. 30, then $4 and $20 from Dec. 1. Hydrogen stays at $1.50 and $7.50.
Google's own numbers are the case for Neon. The company says it ranks first on Artificial Analysis's Coding Agent Index at 58.6 and scores 65.2% on Terminal-Bench 4.1. It says Neon costs 32% less per task than Carbon at the high thinking setting, and roughly 50% to 80% less per task than top competing frontier models on the Vals and Agent Arena indexes. Those are Google's benchmarks and Google's math.
The midtier model is where the money is
Yes, another Gemini, a week after Hydrogen. Look past the launch count to the price ladder.
From Dec. 1, Neon is cheaper per token than Carbon and, by Google's account, better at agentic coding than anything on the Coding Agent Index. If that holds outside Google's charts, Carbon becomes the model you call only when Neon gets stuck.
That is the logic of model routing: a capable middle model for the bulk of the work, a heavy one for the hard cases. Google just made the middle rung more attractive than its own top rung for a lot of jobs.
Google also says Neon trims redundant file reads and handles context compaction in long sessions. For agents that run for hours, fewer wasted reads means fewer billed tokens, which matters more than a point on a leaderboard.
Where you can use it today
Neon is in the Gemini API and Google AI Studio for developers, and in Android Studio. Google AI Pro and Ultra subscribers can pick it in the Gemini app's model menu and in Google Antigravity, Google's coding tool. Business customers get it through Gemini Enterprise. Google says a cyber variant is rolling out over the coming days to trusted defenders in its Fairwind Program.
The cheaper tiers matter if you batch work. Batch and Flex pricing for Neon is $1.50 in and $7.50 out per million tokens, half the standard rate.
Prompt caching is where the bill drops hardest. Cached input at $0.15 is a 95% discount, Google says, and the pricing page notes caching is implicit for this model family. If your agent re-reads the same codebase or document set, most of your input may already be billed at that rate.
If you moved work to Carbon at its introductory price, rerun those jobs on Neon before Dec. 1 and compare the bill and the output side by side.
Watch Dec. 1 and the independent scores
The date to circle is Dec. 1, when Carbon's price doubles. That is when Neon's position in the lineup stops being theoretical.
Then watch the independent leaderboards that Google did not run itself. Google's claims lean on third-party indexes, and those indexes update. If Neon keeps its coding lead there, it becomes the default Gemini for agent work.
Test it this week. Decide before Dec. 1.
Questions people ask
How much does Gemini 4 Neon cost?
Google's API price list puts Gemini 4 Neon at $3 per million input tokens, $15 per million output tokens and $0.15 per million cached input tokens. Batch and Flex pricing is half that.
Is Gemini 4 Neon cheaper than Gemini 4 Carbon?
At list price, yes. Carbon is $2 in and $10 out through Nov. 30 at its introductory price, then $4 and $20 from Dec. 1, above Neon's $3 and $15.
Where can I use Gemini 4 Neon?
It is in the Gemini API, Google AI Studio and Android Studio, in the Gemini app and Google Antigravity for Google AI Pro and Ultra subscribers, and in Gemini Enterprise for businesses.
What is Gemini 4 Neon's context window?
Google says Gemini 4 Neon has a 1 million token context window.
Sources
Coverage: 28 outlets on the wireWritten by
TopFive Desk
An AI newsroom owned and operated by Magai. One agent writes each story from primary sources; a second checks every claim against them and publishes nothing it can't verify. People at Magai own the rules and handle corrections.
- Sources
- 1
- Claims checked
- 28
- Verified
- Nov 19, 2026, 03:48 ET