Changelog
Gemini 4 Neon is available on launch day
Google released Gemini 4 Neon on November 18, and it's in General Input the same day. Every workspace can pick it for specialists, workflows, and chats, on our keys with no setup, and members whose Google plan includes it can run it on their own subscription through Token Tunnel. Google places Neon between Gemini 4 Hydrogen and Gemini 4 Carbon, for complex coding and analysis. Artificial Analysis scores it up to 61 on its Intelligence Index, against 51 for Hydrogen at high. Auto and your default model don't change. Compare it with every other model on the models page.
How it works
- Pricing and window. $3 input and $15 output per million tokens, and $0.15 per million cached input tokens, at one price across its 1M-token window, with image and PDF input.
- Effort. Neon defaults to high, with low and medium to pick from. In our tests it reasoned before answering at every level, on every route.
- Routes. Google Vertex AI serves it first, with Vercel AI Gateway and OpenRouter behind it. When it's overloaded, runs fall back to Claude Opus 5.5, then GPT-6.2 Sol.
- Token Tunnel. Google offers Neon in Antigravity on AI Pro and Ultra plans. The tunnel lists it as soon as your signed-in Google account serves it, under the name Gemini 4 Neon, and plans for 200,000 tokens of context on it until we've measured how much Antigravity accepts.
Also shipped
- Workflows and apps that call our hosted chat completion service can now name Gemini 4 Neon as their model.