Google unveils Gemini 4 Neon; Pro and Ultra subscribers can select it in the app
Image: AI-generated illustration (not an actual photograph)

[DiaryPoint | Reporter Eom Ju-yong] Google unveiled its new AI model, Gemini 4 Neon, on the 18th and began rolling it out to Google AI Pro and Ultra subscribers in the Gemini app, as well as to developers. In the app, users can choose Neon from the model picker.

Consumers can access Neon in the Gemini app with a Pro or Ultra subscription. Developers can try the model in Google AI Studio or connect to it through the paid Gemini API. Neon is also available in Google Antigravity to Pro and Ultra subscribers.

Within the Gemini 4 lineup, Neon is designed for frequent complex tasks, balancing speed with depth of reasoning. Google said, “Carbon handles the toughest, longest-running projects; Hydrogen is for fast, everyday tasks; and Neon strikes a balance between depth and efficiency.”

Previously announced Carbon is geared toward complex, time-consuming projects, while Hydrogen emphasizes quick execution for everyday tasks. Neon is aimed at document and spreadsheet analysis, web research, and coding and knowledge work that involve multiple steps. Google said Neon can handle complex, multistep tasks faster and more efficiently than larger models.

The Gemini API for developers costs $3 per million input tokens and $15 per million output tokens. Cached input costs $0.15 per million tokens, and the context window is 1 million tokens. API usage is billed separately from app subscriptions, based on the number of tokens used.

Google said Neon scored 61 on the Artificial Analysis Intelligence Index with the xhigh reasoning setting, ranking third overall. Whether Neon appears in the app’s model picker may vary depending on an account’s subscription tier and the rollout schedule.