On November 19, 2026 (November 18 in the US), Google released Gemini 4 Neon. It is available in the Gemini API, Google AI Studio, Gemini for business, Google Antigravity and the Gemini app. Antigravity access is limited to paid Google AI Pro and Ultra subscribers; the Gemini app is also Pro/Ultra only. A cyber-defender version for the Fairwind Program opens over the next few days.
Neon is a lighter flagship adapted from Gemini 4 Carbon: Carbon handles the hardest long-horizon projects, Hydrogen handles fast everyday execution, and Neon sits in between, aiming at high-frequency complex work. API pricing is $3 per million input tokens, $0.15 for cached input (95% off) and $15 per million output tokens, with a 1M-token context window and a March 2026 knowledge cutoff.
Google says Neon scores 61 at xhigh on the Artificial Analysis Intelligence Index, third overall behind GPT-6.2 Astra and Claude Fable 5.2, and recorded the lowest fabrication rate of any Gemini 4 model in internal Antigravity testing. It ranks first on the Artificial Analysis Coding Agent Index, scores 65.2% on Terminal-Bench 4.1 and is state-of-the-art on several cyber benchmarks. Those are benchmark results; real-world performance still depends on plan, quota and harness.
For subscribers this is a model update, not a quota reset. Google AI Pro/Ultra users can select Neon in Antigravity; Google has not announced extra resets or reset cards alongside it. Neon has four thinking levels, and Google says handing complex tasks to it can stretch usage limits further in the Gemini app.