Google launches new AI platform AI Edge Gallery for Mac users. Through this, users can run AI models on their own devices without internet connection.

Also Read

80 percent of world’s data centers at climate risk

Bank account can be empty in trap of this app, be careful how

The most popular South Korean content on Netflix after the US

Unlike cloud-based AI services such as ChatGPT, Cloud or Gemini, the locally (on-device) managed AI model operates entirely on the user’s computer. As a result, it is possible to use AI without the Internet and there is no need to send personal information to external servers, which helps in protecting privacy.

Advantages of using native AI
Local AI models are typically smaller than large cloud-based models, but they have several advantages. Users can continue working offline, get faster responses on powerful hardware, and store sensitive data on their own devices.

Until now, Mac users relied mainly on third-party software like Olama or LM Studio to run native AI models.

Journey with five Gemma models
Google’s own five instruction-tuned Gemma models are initially added to the AI ​​Edge Gallery. These are –

JEMA-4-12B-IT
GEMA-4-E2B-IT
JEMA-4-E4B-IT
GEMA-3N-E2B-IT
JEMA-3N-E4B-IT
The app currently focuses only on Google’s Gemma series of models, without supporting models from various developers like Olama or LM Studio.

The new multimodal GEMA 4 12B

At the same time, Google also unveiled a new multimodal AI model called Gemma 4 12B. The company claims that this model of 12 billion parameters is capable of providing equivalent performance to a much larger model of 26 billion parameters.

The model can also run on normal laptops with 16GB RAM. It is capable of text, image and audio processing and can be used for various tasks such as content creation, software development and data analysis.

New AI dictation app
Along with AI Edge Gallery, Google has also launched a new dictation app called AI Edge Eloquent for Mac users.

This free app converts the user’s speech into written text and automatically removes unnecessary words to make the writing clearer and more readable.

Since all processing is done inside the device, privacy of user information is maintained. Users can also customize the output type and add special names, technical terms or industry-specific terms.

The tech sector is turning to on-device AI
According to experts, the initiative is part of a larger trend in the tech sector, which is moving away from reliance on cloud-based infrastructure and focusing on managing AI directly on personal devices.

As hardware capabilities increase, tech companies are investing more in developing on-device AI technologies to increase privacy, respond faster, and reduce internet dependency.

Google’s AI Edge Gallery is already open for macOS. Earlier Android and iOS versions were available. The company is expected to further expand the range of on-device AI tools in the future.