谷歌发布了 EmbeddingGemma 2，这是一款拥有7.4亿参数的开源模型。该模型能够将文本、代码、图像、音频和视频映射到同一个向量空间，且完全可在手机或小型开发板上运行。该模型采用 Apache 2.0 许可证，未经任何安全调优处理；据谷歌称，其在所支持的100种语言中表现并不均衡。
Google has released EmbeddingGemma 2, an open 740-million-parameter model that maps text, code, images, audio and video into a single space and runs entirely on a phone or a small board. It is licensed under Apache 2.0, has had no safety tuning, and Google says performance is not equal across the 100 languages it supports. Google has released an open model that searches text, code, images, audio and video, and runs on a phone.
这款开源模型可处理文本、代码、图像、音频及视频数据，且能在手机上直接运行。在 Pixel 11 Pro 手机上处理文本时，该模型仅需约191MB内存；若需处理全部五种类型的数据，则需约567MB内存。所有数据处理过程均无需将数据传送到外部设备。
It needs about 191MB of memory for text work on a Pixel 11 Pro, or about 567MB for all five. Nothing has to leave the device. EmbeddingGemma 2 maps all of it into one 768-dimension space, Google said in a blog post.
据谷歌在博客中介绍，EmbeddingGemma 2 能将各类数据映射到同一个768维的向量空间。该模型共有7.4亿个参数：其中2.7亿参数专用于文本处理，1.7亿参数用于图像编码，另有3亿参数构成的音频编码器仅在需要时才启用。该模型同样采用 Apache 2.0 许可证。
It has 740 million parameters, of which 270 million are needed for text alone, with a 170-million vision encoder and a 300-million audio encoder loaded only if wanted. The licence is Apache 2.0. An embedding model turns content into numbers so software can find things by meaning rather than by matching words.
嵌入模型的作用是将各类内容转化为数值形式，从而使软件能够基于语义而非单纯的文字匹配来检索信息。通常情况下，这一处理步骤需要在第三方云端服务器上完成。
That is the step that normally happens in somebody else’s cloud. We reported in April that Gemma 4 arrived as four open-weight models under the same licence, running on phones, Raspberry Pi boards and consumer graphics cards.
今年4月曾有报道称，谷歌推出了四款基于相同许可证的开源模型 Gemma 4，它们同样可在手机、树莓派开发板以及普通显卡上运行。EmbeddingGemma 2 正是基于该架构开发而成，并沿用了 Gemma 4 的文本分词器与音频编码器，因此两者协同运行时所需的总内存占用更低。这样一来，便构建出了一个完全无需网络连接的检索系统。
EmbeddingGemma 2 is built on that architecture and shares its text tokeniser and audio encoder, so the two run together in less combined memory. The practical result is a retrieval pipeline with no network in it. Google made the same argument in August when it built an offline translator on an $80 Raspberry Pi, for handling sensitive conversations without sending audio anywhere.
今年8月，谷歌曾在售价仅80美元的树莓派设备上开发出一款离线翻译工具，专门用于处理敏感对话内容，且全程无需上传任何音频数据。将数据保留在本地设备内即可避免数据传输，而数据传输正是引发欧洲各国数据保护争议的主要环节。谷歌此举并非出于慷慨之举，而是旨在确立新的行业标准。
Keeping the data on the device avoids the transfer, which is where most European data protection questions begin. The company is not doing this out of generosity. It is building the default.
谷歌表示，Gemma家族的下载量已经超过了10亿次，其中第一个EmbeddingGemma占了2000万次。新模型在每种模式下拥有8，192个令牌，是第一版的四倍，相当于29张图像，58个视频帧或5.5分钟的音频。
Google says the Gemma family has passed a billion downloads, with the first EmbeddingGemma accounting for 20 million of them. The new model holds 8,192 tokens across every modality, four times the first version, which works out as 29 images, 58 video frames or five and a half minutes of audio.
矢量可以从768维削减到128维，谷歌表示这将存储空间减少了六倍。
Vectors can be cut from 768 dimensions to 128, which Google says reduces storage up to sixfold.
有限制，谷歌列出了它们。该模型没有安全调整，也没有输出调节，而是将缓解措施应用于训练数据。它支持100多种语言，谷歌表示这些语言的性能可能不相同。欧盟有24种官方语言。
There are limits, and Google lists them. The model has had no safety tuning and no output moderation, with the mitigation applied to the training data instead. It supports more than 100 languages, and Google says performance may not be equal across them. The EU has 24 official languages.
欧洲一直表示希望在本地拥有这种能力，在欧洲公司已经拥有的硬件上运行，任何人都可以使用的许可证下，而且它是在山景城建造的。
It is the kind of capability Europe keeps saying it wants to hold locally, running on hardware a European company already owns, under a licence anyone can use, and it was built in Mountain View.