Google · Text models

Gemini 4 Neon launches: Google's mid-tier model for coding and enterprise work

Google released Gemini 4 Neon on Nov 18: GA on the paid Gemini API at $3 / $15 per million tokens, 1M context, in the Gemini app for AI Pro and Ultra.

Google released Gemini 4 Neon on November 18, 2026. Google says it brings “frontier-level intelligence with fast end-to-end task completion” and positions it between Gemini 4 Carbon (the hardest long-horizon projects) and Gemini 4 Hydrogen (fast everyday execution). Google says Neon “balances depth and efficiency” for complex coding, analytical reasoning and enterprise workflows.

What to know

  • Status: the Gemini API release notes list gemini-4-neon as generally available (GA) on the paid tier. No free API tier is mentioned.
  • Price: Google lists $3 per million input tokens, $0.15 per million cached input tokens and $15 per million output tokens, with a 1 million token context window.
  • Thinking levels: low, medium, high (the default) and xhigh, the same four settings as Hydrogen (our thinking level prompts).
  • Where you can use it: the paid Gemini API in Google AI Studio and Android Studio, Google AI Studio for Pro and Ultra subscribers, Google Antigravity for paid Google AI Pro and Ultra subscribers, Gemini for business (API and model dropdown), and the Gemini app model dropdown for Google AI Pro and Ultra subscribers. Google says access for eligible members of its cyber-defender program rolls out over the next few days.
  • Cost per task (Google’s figures): Google says that at its default high level Neon has a 32% lower cost per task than Carbon.
  • Benchmarks (Google’s own figures): 61 on the Artificial Analysis Intelligence Index at xhigh (Hydrogen reached 54 at xhigh), 65.2% on Terminal-Bench 4.1 (Hydrogen: 57%) and 75.8% on Agents’ Last Exam v1.1. Google reports first place on the Artificial Analysis Coding Agent Index (58.6).
  • Agents: Google says Neon can act as an Orchestrator in Antigravity that directs parallel Hydrogen subagents.
  • Safety (Google’s figures): Google says Neon uses the Gemini 4 safety stack with real-time safeguards against cyber and biosecurity misuse, and links a model card.

Why it matters

Neon fills the gap between cheap, fast Hydrogen and expensive, token-hungry Carbon. For multi-file debugging, spreadsheet and PDF analysis or long agent runs where Hydrogen falls short, it is now the next Gemini model to try before Carbon. Benchmarks are Google’s own, so test on your own work. Our updated Gemini 4 model guide compares all four models, and the Gemini page and model tracker have the wider picture.

Source: Google. We summarise in our own words; check the source for the full details.

Related

More AI news

stay in the loop

New guides, prompts and AI model news by email