Google launches Gemini 4 Neon AI model with 1 million-token context window

What happened
Google introduced Gemini 4 Neon, a new AI model, on November 18, 2026. The model is now available to AI Pro and Ultra subscribers through the Gemini app, Google AI Studio, and Google Antigravity. Gemini 4 Neon replaces the previous Argon and 3.1 Pro models and is designed for advanced coding, knowledge work, and scientific tasks. It offers Carbon-level performance at faster speeds and a lower cost.
Key details
- Gemini 4 Neon features a 1 million-token context window.
- The model supports four configurable thinking levels.
- Pricing is set at $3 per million input tokens and $15 per million output tokens.
- Neon leads or ranks highly on multiple benchmarks, including Artificial Analysis Intelligence Index (score: 61 at xhigh), Vals Index (#2, 70.9%), and Agent Arena Index (#2, 1336).
- It offers improved debugging, document processing, and multi-step reasoning capabilities.
Why it matters
AI Pro and Ultra subscribers now have access to a faster, more capable model for tasks in software engineering, analytics, and science, with lower costs per task compared to previous models. Enterprises and developers can handle larger workloads and more complex problems due to the expanded context window and improved reasoning. The change means users must transition from Argon and 3.1 Pro to Neon for continued access to the latest features and performance.
Also covered by 2 sources
- You can use Google’s Spark AI agent, renamed Tasks, for free now – with limits· ZDNet
- Google’s Spark AI agent becomes ‘Tasks’ — and now it’s free· The Verge
- What Is Gemini? Everything You Should Know About Google’s AI Tool· CNET
- Google Antigravity Rolls Out Orchestrator for Dynamic Model Handoffs· Unite.AI
- Google’s latest Search upgrade promises better answers to your most complicated questions· 16 sources
- Google is letting staff test a 'fast' new Gemini AI model named Hydrogen· Business Insider


