Google DeepMind released Gemini 4 Neon on November 18, a mid-tier model for coding and agent work that sits between Hydrogen and Carbon in the Gemini 4 family. It has a context window of about 1 million tokens, shared across input, thinking and output.
Pricing is $3 per million input tokens and $15 per million output tokens, with cached input at $0.15. Google says Neon costs about 32% less per task than Gemini 4 Carbon at high thinking. Against Carbon's introductory $2/$10 rate, which runs through November 30, Neon is 50% more expensive per token; against Carbon's $4/$20 rate from December 1, it is 25% cheaper.
The model is available through Google's API and eligible Gemini subscriptions. Its launch post gives no absolute completion times, and the per-task savings are Google's own claim.
