AI NEWS · EDITORIAL SUMMARY
Google introduces Gemini 4 Neon for complex coding and analysis
The new model targets a balance of depth, speed and cost for professional work, with paid-plan and API access.

Why it matters: It combines document analysis, coding and agent workflows within the Gemini model family.
Where Neon fits
Google adds Neon to the Gemini 4 family for frequent workloads needing advanced reasoning. The company positions it between Carbon, for difficult long-running projects, and Hydrogen, for fast everyday execution.
Capabilities in one model
The announcement highlights spreadsheets, PDFs, charts, audio and video within a one-million-token context. Coding capabilities include debugging and refactoring across multiple files, as well as coordinating agents in Antigravity.
Access and terms
Rollout includes Gemini API and AI Studio, Gemini for business, and Pro and Ultra plans in the Gemini app and Antigravity. API pricing starts at $3 per million input tokens and $15 per million output tokens; cached input has a separate rate. Access depends on the channel, account and its limits.
More reasoning control
Neon provides four thinking levels. Google reports a lower cost per task than Carbon at the default high level in the cited evaluations; this does not mean a general reduction in the price per token. The xhigh level applies deeper reasoning to harder problems.
CONTINUE WITH THE ORIGINAL
Read the full announcement
This is Serconet’s own editorial summary. The original announcement provides the full details and conditions.
Read the original · Google · English (opens in a new tab)