Technology & AI

Google releases Gemini 4 Neon AI model with million-token context

The launch follows several new Gemini models from Google in about a month, with the Neon model priced above earlier versions.

By Standard Topic. AI-generated from the reporting listed under Coverage. Published . Updated .

What happened

Google released Gemini 4 Neon, a new model built for software development, knowledge work and science. It handles text, images, video, audio and PDFs with a context window of one million tokens. It is priced at $3 per million input tokens and $15 per million output tokens, rolling out to AI Pro and Ultra subscribers in the Gemini app, Antigravity and the Gemini API.

Why it matters

Google described Neon as sitting between its Carbon and Hydrogen models, with deeper reasoning and faster performance than its flagship, at lower pricing. Reports also questioned how much coding-agent performance comes from the underlying model, since a coding benchmark score rose from 12.8% to 39.0% when run inside Antigravity.

Timeline

  1. Nov 18, 4:03 PM ET: Google released Gemini 4 Neon, its fourth new Gemini model in a month, with a context window of one million tokens.
  2. Nov 18, 4:43 PM ET: Gemini 4 Neon ranked second on Arena.ai's Agent Arena Index, trailing only Claude Fable 5.2.
  3. Nov 18, 7:45 PM ET: A coding benchmark score for Gemini 4 Neon rose from 12.8% to 39.0% when the same model ran inside Antigravity.

Coverage

More from Technology & AI

All Technology & AI news · Technology & AI story archive