Google has launched Gemini 4 Neon with powerful programming capabilities

The new AI model combines a high level of reasoning with the ability to perform complex tasks efficiently.

0

Google has unveiled a new artificial intelligence model, Gemini 4 Neon, designed for complex engineering, analytical and scientific tasks. The model supports a context window of one million tokens and delivers high performance in programming, research and workflow automation.

Key points at a glance:

  • Google has unveiled Gemini 4 Neon for complex professional tasks.
  • The new model supports a context window of one million tokens.
  • Gemini 4 Neon came third in the Artificial Analysis ranking.
  • The model is available to developers and subscribers to Google AI Pro and Ultra.
Add UkrMedia to your Google sources Get more of our news in the recommendations.

What the Gemini 4 Neon can do

Google announced the new model on 18 November. It is designed for work on large-scale software projects, document analysis, financial data processing and the execution of multi-stage tasks.

The company had previously unveiled Gemini 4 Hydrogen, which has also been integrated into the Google search engine.

Gemini 4 Neon is capable of working with various types of information, including spreadsheets, PDF documents, images, audio and video.

According to Google, the model can analyse programme code independently, identify errors and carry out complex tasks that require coordination between several tools.

Test results for the new model

Google has published the results of the Gemini 4 Neon in a series of international tests.

Test Result
Artificial Intelligence Analysis Index 61 marks
Vals Index 70,9%
Terminal-Bench 4.1 65,2%
Terminal-Bench Science 70,5%
FrontierCode 2.0 46,2%
HLE-Diamond 71,3%

In the Artificial Analysis Coding Agent Index ranking, the new model took first place according to data provided by Google at the time of the presentation.

How much does the Gemini 4 Neon cost?

The company has set the following API rates for developers:

  • $3 per million incoming tokens;
  • $0.15 per million cached incoming tokens;
  • $15 per million initial tokens.

Gemini 4 Neon is gradually becoming available via Google AI Studio, Android Studio, Google Antigravity and the Gemini enterprise platform.

In the standard Gemini app, the new model is available to Google AI Pro and Google AI Ultra subscribers.

WRITE A REPLY

enter your comment!
enter your name here