谷歌向低内存安卓设备开放Gemini应用 推出Gemini 4 Neon主攻复杂编程100万token上下文Google Brings Gemini App to Low-Memory Android Devices, Launches Gemini 4 Neon for Complex Coding with 1M-Token Context


AI free token/AI免费积分

谷歌向低内存安卓设备开放Gemini应用 推出Gemini 4 Neon主攻复杂编程100万token上下文

本文来源:https://news.qq.com/rain/a/20261120A01UAR00
若有版权问题,请来信告知:13951431913#139.com 或在右方提交评论。

2026-11-20 01:49 发布于 广东 观点地产新媒体官方账号

观点网讯: 11月19日消息,谷歌发布博文,宣布面向Android Go设备开放Gemini应用,经重新设计的应用可在内存低至2GB的手机上流畅运行。

另据消息,谷歌宣布推出AI模型Gemini 4 Neon,主攻复杂编程。该模型支持100万token上下文窗口,可跨文件排查错误、重构代码,还能在Antigravity环境中协调多个AI子智能体协作。

Neon在Artificial Analysis编程智能体指数中以58.6分排名第一;默认high档位下单任务成本比同系列Carbon模型低32%。API输入每百万词元3美元、输出15美元。

Google Brings Gemini App to Low-Memory Android Devices, Launches Gemini 4 Neon for Complex Coding with 1M-Token Context

本文来源:https://news.qq.com/rain/a/20261120A01UAR00
若有版权问题,请来信告知:13951431913#139.com 或在右方提交评论。

Google opens the Gemini app to low-memory Android devices and launches a new model, Gemini 4 Neon

Posted on 2026-11-20 at 01:49 from Guangdong, by the Guandian Property New Media official account

Guandian.com reports: On November 19, Google published a blog post announcing that it is opening the Gemini app to Android Go devices, with the redesigned app running smoothly on phones with as little as 2GB of memory.

According to another report, Google announced the launch of the AI model Gemini 4 Neon, focused on complex programming. The model supports a 1 million token context window, can debug errors and refactor code across files, and can coordinate multiple AI sub-agents to collaborate within the Antigravity environment.

Neon ranked first with a score of 58.6 on the Artificial Analysis coding agent index; at the default 'high' setting, the per-task cost is 32% lower than the Carbon model in the same series. The API costs $3 per million input tokens and $15 per million output tokens.

暂无评论,快来发表第一条评论吧!No comments yet. Be the first!

📮 需求咨询📮 Inquiry