AI Digest: Grok 4.5, AI Chip DeepSeek, and the Largest AI Data Center in Central Asia

xAI Unveils New AI Model Grok 4.5
xAI has announced its new Grok 4.5 model, designed for programming, complex reasoning, and multi-step task execution. According to developers, the model is faster, more accurate, and better at writing code, analyzing data, and performing mathematical calculations.

DeepSeek Develops Its Own AI Chip
Chinese startup DeepSeek has begun developing its own artificial intelligence processor. The company aims to reduce dependence on foreign accelerator suppliers and create its own hardware platform for training and running large language models.

One of Central Asia's Largest AI Data Centers to Be Built in Uzbekistan
DataVolt will build one of the largest AI-ready data centers in Central Asia in Uzbekistan. The new facility will be part of the country's digital infrastructure development and will be designed to support high-performance computing and artificial intelligence projects.

zypl.ai to Help Integrate Artificial Intelligence into Tajik Schools
Tajik AI startup zypl.ai will participate in integrating artificial intelligence into the country's education system. The company will help integrate AI tools into schools to enhance digital literacy and introduce students to modern technologies.

AI Cluster to Be Created in Kazakhstan with Participation from Citi and Freedom Holding Corp.
A new AI cluster has been announced in Kazakhstan, involving international corporation Citi and Freedom Holding Corp. The project aims to develop the artificial intelligence ecosystem, support innovative companies, and attract international partners.

Kaspi.kz Launches AI Assistant Kasper
Kaspi.kz has introduced Kasper, a new AI assistant designed to help users find products faster, compare offers, and receive personalized recommendations while shopping within the company's ecosystem.

Google Unveils Faster and More Accessible Image Generator Nano Banana 2 Lite
Google has released Nano Banana 2 Lite, a faster and more accessible image generation model. The solution is aimed at developers and users who require high speed while maintaining generation quality.

SoftBank Begins Mass Production of AI Robots
Japanese SoftBank has announced the start of mass production of robots with artificial intelligence. The new generation of humanoid robots is intended for use in industry, logistics, and the service sector.

Google Launches AI Agents in Search
Google has begun gradually rolling out AI agents in Google Search—a new feature that allows artificial intelligence to independently track information on the internet, search for products, monitor price changes, and perform other tasks without constant user queries. Simultaneously, the company announced that AI Mode is already used by over 1 billion people per month, powered by the new Gemini 3.5 Flash model.

Meta Unveils AI Image Generator Muse Image
Meta has launched the free image generator Muse Image, integrated into Meta AI, Instagram Stories, and WhatsApp. The service allows users to create and edit images based on text prompts, but it has raised privacy concerns: users can use photos from public Instagram accounts to generate new images without prior consent from the owners.

Codex Gains AI Video Editor
ChatCut has released a plugin for Codex featuring a built-in AI video editor. It allows users to edit videos directly within the app without switching between different programs. Users can control editing via text commands, and existing Codex subscribers will receive the plugin for free.

Google Unveils AI Model SensorFM for Health Monitoring
The AI model SensorFM analyzes data from Fitbit and Pixel Watch, including heart rate, skin temperature, blood oxygen levels, and physical activity, helping to detect signs of cardiovascular, metabolic, and other diseases. The model can generate personalized health recommendations.