Google just launched Gemini 4 Neon, its newest AI model, on November 19.
Designed for tough coding and reasoning tasks, Neon is now available to AI Pro subscribers and officially replaces the older Argon model in the Gemini app.
Neon supports 1-million-token context window
Neon can handle a massive 1-million-token context window, helping it ace analytical and financial benchmarks.
It's priced at $3 per million input tokens and $15 per million output tokens, making it pretty affordable for power users.
For those with AI Pro or Ultra plans, Neon is also available in Google Antigravity for paid AI Pro or Ultra subscribers, slotting between Hydrogen and Carbon models, so you get more flexibility for complex projects.