Google's New Gemini 4 Neon Model Is out, and All Its Scores Come From Google

Google began rolling out a new artificial-intelligence model, Gemini 4 Neon, on Nov. 18, 2026. The company said it is available in the Gemini API in Google AI Studio, in Gemini for business, and for paid Google AI Pro and Ultra subscribers in the Gemini app and in Google Antigravity, its coding tool.
Neon sits in the middle of Google's current model line. The announcement by Koray Kavukcuoglu, a senior vice president at Google DeepMind, describes it as built for complex coding, analytical reasoning and production-grade enterprise work. The company keeps the larger Gemini 4 Carbon for the hardest long-running projects and the faster Gemini 4 Hydrogen for everyday work.
Google published the model's terms: $3 per million input tokens, $0.15 per million cached input tokens and $15 per million output tokens, with a 1 million-token context window — the amount of material the model can take in at once. Google's post lists four thinking levels, settings that trade how much reasoning the model does against how many tokens a job costs.
Every capability figure in the announcement is Google's own. On Google's account, Neon placed third overall with a score of 61 on the Artificial Analysis Intelligence Index at its highest thinking level, behind larger and slower models; the company also reports placements on the Vals Index and on Arena's Agent Arena Index. No independent evaluation of those claims has been published.
Google said a version for cybersecurity defenders, Gemini 4 Neon for cyber, will reach members of its Fairwind Program in the days after the release.
