News
ChatGPT received Study Mode for user learning mode
OpenAI launched Study Mode in ChatGPT, promoting independent learning through interactive hints, tests, and response analysis.
Hunyuan World generative model creates 3D scenes for VR and games
Tencent has opened access to Hunyuan World Model 1.0, which generates three-dimensional virtual scenes based on text and images.
Grok will receive infinite image generation and audio videos
xAI announced a major update for Grok with a new companion and image and video creation feature directly in the…
Harmonic App with Aristotle Promises Accurate Mathematical Answers
Harmonic released a beta version of a chatbot with the mathematical AI model Aristotle, providing formally verified answers.
Gemini 2.5 receives “conversational image segmentation” feature
Google introduced a feature for Gemini 2.5 that highlights parts of images using user text queries.
The new AI tool Figma Make is now available to all users
Figma opened the Make tool to everyone, added image generation and editing, and introduced an AI credit system
Alibaba releases four open Qwen3 models for business
Alibaba has introduced new Qwen3 models specializing in complex thinking tasks, programming, and translation, with a free commercial license.
New AI Technology Provides 100 Times Faster Reasoning
The new HRM model from Sapient Intelligence shows better accuracy in complex tasks than large language models.
Midjourney now allows setting initial and final images for video
Midjourney has added features for creating videos with image transitions and improved functionality in Discord
Google Introduces Opal for Rapid AI App Development
Google launches Opal — a convenient service for developing, testing, and sharing applications with support from Gemini, Veo, and Imagen…
Google Photos added new tools for animating and stylizing photos
Google Photos adds tools for creating videos from photos, image stylization, and digital content marking.
Microsoft prepares its Copilot service for the launch of GPT-5
Copilot will receive a Smart mode that independently selects the optimal approach to user queries after the release of GPT-5
