Urgent.News

What's breaking now, across thousands of outlets.

AI

Introducing Gemini 4 Neon

Today, we’re introducing Gemini 4 Neon

Introducing Gemini 4 Neon

Google has unveiled Gemini 4 Neon, a state-of-the-art AI model designed to tackle complex software engineering, knowledge work, and science workflows. With a score of 61 on the Artificial Analysis Intelligence Index, Gemini 4 Neon ranks just behind larger, slower models. However, it shines in areas such as multi-step instruction following and steerability, ranking #2 on the new Agent Arena Index.

Priced at $3 per million input tokens and $15 per million output tokens, Gemini 4 Neon offers a significant 95% discount on input tokens compared to previous models. The model boasts an impressive 1 million-token context window, allowing it to synthesize data from various sources, including spreadsheets, PDFs, charts, audio, and video, into structured deliverables.

Gemini 4 Neon has consistently outperformed comparable models, ranking #1 overall on the Artificial Analysis Coding Agent Index (58.6) and achieving a 65.2% score on Terminal-Bench 4.1. It is also leading in SWE-sweep mini-SWE harness (12.8%) and more than triples to 39.0% in Google Antigravity, showcasing its potential for long-running agent workflows.

Beyond its coding capabilities, Gemini 4 Neon excels in scientific and mathematical reasoning, with a 73.5% score on Chartography for reading dense professional charts and a 75.8% score on Last Exam v1.1. In the realm of cybersecurity, Neon for cyber delivers frontier-level performance in vulnerability finding, validating, and patching, all while being faster and more cost-efficient than larger models. It is set to roll out to trusted defenders in the Fairwind Program.

Written by urgent.news from Google Blog's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.

Also reported by 1 other outlet

Every outlet that reported this, and when →

Read the original at blog.google →

More in AI

More from Wednesday 18 November →