llm-gemini 0.37

Google has released Gemini 4 Hydrogen, the lowest‑cost variant in the Gemini 4 family. According to the published pricing table, Hydrogen charges $1.50 per million input tokens, $0.075 per cached token, and $7.50 per output token. This represents a 25 % discount compared to its siblings Carbon and Argon before December 1, and a 75 % discount thereafter. Carbon and Argon share the same pricing—$2 per million input tokens and $10 per output token—and both will double their rates starting December 1. The difference between them is that Argon delivers faster responses and consumes fewer tokens per turn.

To illustrate the trade‑off between quality and cost, the author generated images of pelicans on bicycles with all three models and three levels of “thinking.” The results show that Carbon and Argon produce detailed, rich scenes, while Hydrogen produces much simpler illustrations, albeit at a fraction of the time and expense.

Google positions Hydrogen as a daily‑use model suitable for coding, testing, and verification tasks, and has designated it as the free model that will support AI features in Google Search. This announcement is significant because it clarifies Google’s pricing strategy and helps developers choose the model that best balances cost and capability for their applications.

Read the original article on Simon Willison

This summary is an informational synthesis produced by dataqbs.com. All rights to the original content belong to its author and the cited media outlet. We act solely as curators of technology news and claim no authorship.