Urgent.News

What's breaking now, across thousands of outlets.

AI

Google’s Gemini 4 Neon tripled its coding score without changing the model

Google’s new Gemini 4 Neon resolved 12.8% of the SWE-sweep coding benchmark using the mini-SWE harness, but that score jumped The post Google’s Gemini 4 Neon tripled its coding score without changing the model appeared first on The New Stack .

Google’s Gemini 4 Neon tripled its coding score without changing the model

Google's Gemini 4 Neon model has shown a significant increase in coding performance, tripling its score on the SWE-sweep benchmark when run within the Antigravity harness, jumping from 12.8% to 39.0%. This substantial improvement has raised questions about the extent to which the underlying model contributes to coding-agent performance and how much is influenced by the harness managing its tools, context, and execution.

The company's testing did not isolate these factors, leaving uncertainty about which changes contributed most to the SWE-sweep improvement.

Brief written by urgent.news from The New Stack's own syndicated text. Machine-written — may contain errors; check the original before relying on it.

This story

This is one outlet's version. Read the fullest account.

Every outlet that reported this, and when →

Read the original at thenewstack.io →

More in AI

More from Thursday 19 November →