这些模型现在能在保持内容安全标准的同时，改进了模型的输出风格，旨在实现更精确、更经济高效的使用。此外，谷歌还推出了 Gemini-1.5-Flash-8B-Exp-0924 升级版，进一步增强了文本和多模态应用。

谷歌公司近日宣布向安卓开发者开放 Gemini Nano 轻量 AI 模型的试用。该模型专为设备端轻量级任务设计，适用于性能较低的手机和平板电脑。开发者可以使用谷歌提供的 AI Edge SDK 将该模型整合到应用中。

Google might put an AI button on the lock screen. It's not clear what it'll do, but it could be related to its Gemini AI.

The Gemini feature will reportedly let users highlight a section of the image and add a text prompt to change the details.

Google announced Thursday that it is expanding its Gemini Live feature to 40-plus new languages and connecting Gemini AI to ...

Google has now made a significant move toward the broader adoption of Gemini Nano by allowing Android developers to ...

Google says that Gemini is subject to the same enterprise terms as other core Workspace services like Gmail and Docs, and won ...

The integration of the Gemini Nano model into the Dimensity 9400 chipset means that any smartphone with this chip will be ...

自发布以来，Gemini 1.5 Flash 模型凭借其速度和成本效益迅速受到开发人员的青睐。 今年 8 月，Google发布了 Gemini 1.5 Flash 8B，这是一个新的实验性人工智能模型，只使用 80 ...

10 月 8 日消息，联发科官宣新一代处理器已针对 Google 大语言模型（LLM）Gemini Nano 多模态进行优化。此外新一代天玑 9400 处理器还搭载第 8 代 MediaTek AI ...

Google said on Thursday it is now testing out generative AI-powered video search in India. This feature enables users to take ...

Google在搭载 Tensor G4 芯片组的 Pixel 9 系列上推出了 Gemini Nano。 它可以帮助智能手机在设备上执行人工智能任务，而无需通过互联网连接到服务器。 Gemini Nano ...