Welcome to the "AI Daily" column! This is your guide to exploring the world of artificial intelligence every day. Every day, we present you with the latest content in the AI field, focusing on developers to help you understand technology trends and innovative AI product applications.
Fresh AI products click to learn more: https://app.aibase.com/zh
1. Generative Video Evolves Again! Veo 3.1 Brings Character Consistency and Multi-Scene Story Generation Features
The Veo 3.1 version brings multiple important improvements for content creators and video producers, including increased video duration, enhanced character consistency, improved output quality, and a new "Multi-shot Multi-prompt" feature. These updates significantly enhance creation efficiency and professional visual appeal.
【AiBase Summary:】
🎥 Video length extended to one minute, meeting longer narrative needs.
👥 Enhanced character consistency, making characters more coherent across different scenes.
🎬 Multi-shot multi-prompt feature, generating multi-scene story scripts with one click.
2. Ant Group Launches Trillion-Parameter Language Model Ling-1T, Setting New Industry Standards in Reasoning Speed and Capabilities
Ant Group's Ling-1T language model has shown excellent performance in reasoning speed and capabilities, marking an important breakthrough in domestic open-source models.
【AiBase Summary:】
🧠 Ling-1T uses a "pre-training + post-training" method to improve reasoning capabilities.
🧮 Excellent performance in math and logic reasoning tasks, solving complex problems.
💻 Supports code generation and travel route planning, with wide-ranging applications.
More details: https://huggingface.co/inclusionAI/Ling-1T
3. xAI Launches Video Generation Model Imagine v0.9, Entering the "Movie-Level One-Click Generation" Era
xAI's Imagine v0.9 video generation model has made significant breakthroughs in visual quality, natural motion, and audio generation, marking a new stage in AI video creation.
【AiBase Summary:】
🎥 Visual quality improved, with higher resolution and richer details, approaching professional movie-level standards.
🔄 Motion generation is more smooth and natural, with seamless transitions between actions.
🎵 Enhanced audio capabilities, allowing simultaneous generation of background music, dialogue, and singing.
4. SoftBank Acquires ABB Group's Robotics Business for $5.375 Billion, Focusing on Physical AI
SoftBank Group announced it will acquire ABB Group's robotics business for $5.375 billion, with the transaction expected to be completed in 2026. The acquisition aims to promote the development of physical AI and strengthen SoftBank's investment strategy in the robotics sector.
【AiBase Summary:】
💰 SoftBank acquires ABB's robotics business for $5.375 billion, expected to complete in 2026.
🤖 ABB's robotics division has 7,000 employees and generated $2.3 billion in revenue in 2024.
🚀 SoftBank has identified robotics as one of its four key investment areas, aiming to drive the development of physical AI.
5. Vercel v0 Tool Launches New Image Editing Feature: Easy Respawn and Prompt Optimization in Design Mode
Vercel's AI-driven UI generation platform v0 launched a new image editing feature, allowing users to directly edit image elements in design mode, simplifying the iteration process for designers and developers.
【AiBase Summary:】
✅ Design mode supports interactive image editing, improving work efficiency.
🔄 Offers 'Respawn' and 'Prompt Editing' features for quick image adjustments.
🖼️ No need to switch to external tools for image editing, enhancing creative flow.
6. OpenAI Sora2 Shocking Release: First Day Installations Match ChatGPT Launch
OpenAI's Sora2 video model quickly gained attention after its release, with installation numbers surging to third place on the US App Store and surpassing other mainstream AI apps in a short time. However, this technology also raised ethical issues, such as the misuse of deepfake videos.
【AiBase Summary:】
🎥 The release of the Sora2 model caused Sora app to rapidly rise in the App Store, ranking third.
⚠️ The technology has sparked ethical controversies, including the use of AI to generate deepfake videos of deceased celebrities.
📈 Sora's download numbers reached levels similar to ChatGPT on its launch day, showing its huge market potential.
7. Sora2 Free Party! Lovart Unlocks Watermark-Free AI Videos with One Click, 5 Days 5+1 Pro Generation, Pro Members Unlimited Usage
The article introduces the Lovart platform integrating OpenAI's latest Sora2 model and launching a limited-time free promotion, allowing global users to experience watermark-free Sora2 standard and Pro-level generation functions. It also mentions that Pro or Ultimate members can enjoy unlimited access to Sora2, as well as the technical highlights of Sora2 in video generation.
【AiBase Summary:】
🔥 Lovart integrates Sora2 model, providing watermark-free video generation services.
💡 Pro members can enjoy unlimited Sora2 generation, supporting various styles of creation.
🚀 Sora2 has high-resolution and long-duration video generation capabilities, improving content creation efficiency.
8. ChatGPT Receives Major Upgrade! Application Ecosystem Goes Live, Opening a New Era of Intelligent Conversations
At the OpenAI DevDay 2025 conference, OpenAI announced that ChatGPT will introduce a new "Apps" system, allowing users to directly use third-party services within conversations. This update marks ChatGPT's evolution from a single conversation tool into a rich application platform, offering new opportunities for users and developers.
【AiBase Summary:】
🚀 ChatGPT introduces a new "Apps" system, enabling seamless integration with third-party services.
🌐 Users can directly call services like Spotify and Expedia within chats, without switching interfaces.
🛠️ OpenAI released the Apps SDK, allowing developers to build ChatGPT apps based on the Model Context Protocol.
