Google's Gemini 4 Neon lands between Carbon and Hydrogen, at $3 per million input tokens
Sundar Pichai announced the model on X. Google says Neon ranks third on the Artificial Analysis Intelligence Index and costs 32% less per task than Carbon at its default setting.
Google has added a third model to its Gemini 4 line. Chief executive Sundar Pichai posted on X on Wednesday afternoon that Gemini 4 Neon brings “frontier-level intelligence with fast end-to-end task completion across complex software engineering, knowledge work, and science workflows”, and that it “begins rolling out today”.
Where Neon fits
In its launch post, Google places Neon between its two earlier Gemini 4 models: Carbon “tackles your hardest, long-horizon projects”, Hydrogen “handles fast everyday execution”, and Neon “balances depth and efficiency”. We covered Gemini 4 Carbon and Gemini 4 Hydrogen when they arrived.
Background
Earlier stories on this thread.
What happened next
Later stories on this thread.
Price and context window
Google lists Neon at $3 per million input tokens and $15 per million output tokens. Cached input costs $0.15 per million tokens, which Google describes as a 95% discount. The context window is 1 million tokens. At its default “high” thinking level, Google says Neon has a 32% lower cost per task than Carbon. Neon has four thinking levels in total.
What Google claims on benchmarks
These are Google’s own figures and its picks of third-party leaderboards:
- Third overall on the Artificial Analysis Intelligence Index, with a score of 61 at its “xhigh” thinking level, and first on the Artificial Analysis Coding Agent Index (58.6).
- Second on the Vals Index at 70.9%, and first on the Legal Research Bench and Tax Agent Bench.
- 65.2% on Terminal-Bench 4.1 and 39.0% on SWE-sweep when run inside Google Antigravity, against 12.8% on the mini-SWE harness.
- Second on Arena’s new Agent Arena Index with a score of 1336.
Google also says Neon recorded its lowest fabrication rate yet in internal Antigravity testing. That is an internal measure, and Google did not publish the underlying numbers.
Who can use it now
According to Google, Neon is rolling out in the paid tier of the Gemini API in Google AI Studio and Android Studio, in Gemini for business, to paid Google AI Pro and Ultra subscribers in Google Antigravity, and to Pro and Ultra subscribers in the Gemini app. A cybersecurity version, Neon for cyber, will reach members of Google’s Fairwind Program over the next few days.
Our read
Analysis: Neon is priced and pitched as the model most developers would run by default, with Carbon held back for the hardest jobs. The benchmark wins are Google’s selection, so the useful test is cost per finished task in your own workload, which the paid API tier now makes possible.
This report is not legal, financial, or health advice. See the disclaimer.
Primary sources
- Sundar Pichai on X: Gemini 4 Neon begins rolling out today Sundar Pichai on X
- Introducing Gemini 4 Neon Google