Gemini 4 Neon Is Google's New Mid-Tier Model — Faster Than Carbon, Cheaper Too
Google keeps shipping. Less than two months after the Gemini 4 family debuted, there is a fourth member: Gemini 4 Neon, announced November 18 by DeepMind's Koray Kavukcuoglu. It lands right between Carbon and Hydrogen, a
Google keeps shipping. Less than two months after the Gemini 4 family debuted, there is a fourth member: Gemini 4 Neon, announced November 18 by DeepMind's Koray Kavukcuoglu. It lands right between Carbon and Hydrogen, and it is aimed squarely at developers who want strong reasoning without the wait.
Carbon handles the hardest long-horizon work. Hydrogen is the fast, cheap everyday option. Neon takes the middle ground: high-frequency complex coding, analytical reasoning, and enterprise workflows. Google calls it "frontier-level intelligence with fast end-to-end task completion," and the early numbers back up the positioning.
On the Artificial Analysis Intelligence Index, Neon scores 61 at its highest "xhigh" thinking level — third overall, behind GPT-6.2 Astra and Claude Fable 5.2, and a point ahead of Carbon. In coding, it scores 59 on the Coding Agent Index through Antigravity CLI, three points ahead of Carbon.
Speed is where it pulls away. Artificial Analysis measured roughly 110 output tokens per second, close to double Carbon's 57. A typical task runs about $6.44 versus $7.76 on Carbon, and finishes in around 21 minutes instead of 36. One wrinkle: Neon uses about 15% more output tokens per task, so the saving comes from cheaper tokens rather than fewer of them.
Pricing is $3 per million input tokens and $15 per million output tokens, with cached input at $0.15 — a 95% discount. That is 25% below Carbon's standard $4/$20 rate, and it matters from December 1, when Carbon's introductory $2/$10 pricing ends. From that date, Neon is cheaper per token than Carbon as well as faster. It carries a 1 million-token context window and four thinking levels (low, medium, high, xhigh) so teams can trade intelligence against cost.
There is no free tier. Neon is rolling out to Google AI Pro and Ultra subscribers in the Gemini app, where it replaces Argon in the model picker, and to developers through the Gemini API, AI Studio, and Antigravity. Google is also preparing a cybersecurity version aimed at finding vulnerabilities and fixing code.
For most developers, Neon looks like the practical pick of the family. Carbon stays the choice for the hardest single problems; Hydrogen stays the cheapest quick answer. Neon is the daily driver for repo-scale coding and long-running agent work.
Read the full breakdown — including the benchmark details and what changes on December 1 — on Rohan Tech Official: https://rohantechofficial.com/tech-news/google-gemini-4-neon/
Originally published by Dev.to AI. Aggregated on AIWithGhost for educational purposes — full credit and traffic to the original publisher.