Model release · Google
Google releases Gemini 4 Neon across the Gemini API, the Gemini app, Antigravity, Cursor, GitHub Copilot and OpenRouter
- Launched
Google's new frontier model sits between Gemini 4 Carbon and Gemini 4 Hydrogen and targets complex software engineering, knowledge work and science. It offers four thinking levels (low, medium, default high and xhigh), a 1M-token context window and output up to 1M tokens. Google says it scores 61 on the Artificial Analysis Intelligence Index at xhigh, ranking third, and Antigravity says it ranks first on the AA Coding Agent Index. Cursor reports 63.7% on CursorBench 4.1 at default effort, up from 57.3% for Hydrogen. GitHub says it comes close to Carbon in quality at a lower token cost per task and responds faster.
Availability
Generally available in the Gemini API as gemini-4-neon at $3 per million input tokens and $15 per million output tokens on the paid tier, through Google AI Studio, Android Studio and Gemini for business. Google AI Pro and Ultra subscribers can pick 4 Neon (Deeper analysis) in the Gemini app for Chat and Tasks, and paid subscribers get it in Google Antigravity across the desktop app, CLI and IDE extensions. Eligible Fairwind Program members get it later. It is also live in Cursor at the same $3/$15 price, generally available in GitHub Copilot (app, CLI and VS Code) and on OpenRouter. Antigravity's official models docs list Neon for paid Google AI Pro or trials bundled with eligible devices, so some trial Pro plans may not get it; actual access varies by account.