AI Daily: Medeo AI New Video Agent Launches: OpenAI Officially Releases GPT-5.2; DINGTalk AI Features Get Major Upgrades

Welcome to the "AI Daily" section! This is your guide to exploring the world of artificial intelligence every day. Each day, we present you with the latest content in the AI field, focusing on developers to help you understand technical trends and innovative AI product applications.

Fresh AI products Click to learn more:https://app.aibase.com/zh

1、Medeo AI New Video Agent Shakes Up the Market: One-Click Script Rewriting with Natural Language, Making Content Creation Accessible to All!

Medeo AI's new video agent introduces an intelligent agent architecture and natural language editing features, enabling dynamic interaction and efficient iteration in video creation, significantly lowering the barrier to entry and improving output quality.

AiBase Highlights:

🧠 Supports complex prompts and real-time natural language modifications, enhancing flexibility in video creation.

🔄 Enables infinite iterative editing, reducing production cycles by over 50%.

🌐 Multimodal integration and intelligent asset libraries optimize output quality and cover multilingual markets.

More details: https://www.medeo.app/

2、DingTalk 8.1.10 Launches Major Upgrade: Bringing AI Into Every Message, Meeting, and Recording Device

DingTalk 8.1.10 integrates AI deeply into workplace communication through three core functions, enhancing work efficiency and cross-language collaboration experiences.

AiBase Highlights:

🧠 AI Smart Reply: Intelligently generates multiple reply suggestions, improving communication efficiency.

🎙️ AI Listening and Q&A: Upgrades meeting notes into interactive collaborative partners, enhancing information processing capabilities.

🌐 DingTalk A1 Hardware Upgrade: Supports real-time multilingual translation, aiding international collaboration.

3、OpenAI Releases GPT-5.2 at Midnight: Finish a Presentation in Three Minutes, Giving Workers an Extra 10 Hours Weekly for “Slacking Off”

At 2 AM, OpenAI released its latest model, GPT-5.2, into production without a press event, showcasing its powerful features through a 12-minute demo video. GPT-5.2 excels in multiple high-frequency scenarios such as coding, table creation, and presentation making, with significantly reduced latency and improved logical consistency test scores. Additionally, OpenAI announced an adult mode, supporting private deployment in high-risk areas, and emphasized compliance.

AiBase Highlights:

🚀 Latency decreases by 42%, 37%, and 51% respectively in coding, table creation, and presentation making.

💰 OpenAI’s pricing strategy is highly competitive, with a 3000-word industry report costing as low as 0.8 cents.

🔒 The adult mode supports private deployment in high-risk fields such as healthcare, law, and finance, meeting dual compliance requirements from the EU AI Act and China's deep synthesis regulations.

4、Disney Invests $1 Billion for AI Tickets: Mickey Mouse, Iron Man, and Darth Vader All Join OpenAI Sora

Disney has reached a strategic partnership with OpenAI, investing $1 billion for AI technology application, promoting the innovative use of IP content in generative AI, marking the first time a traditional entertainment giant has integrated its core IP into an AI platform.

AiBase Highlights:

✅ Disney signed a three-year strategic agreement with OpenAI, becoming the first “main content licensing partner” on the Sora short video generation platform.

✅ Disney invested $1 billion in cash and received additional equity subscription rights, while integrating the ChatGPT API to develop next-generation interactive content.

✅ Both parties plan to launch features in early 2026, including official selected Sora fan-generated videos available for Disney+ subscribers.

5、Google Gemini Deep Research Gets a Major Upgrade: For the First Time, Developers Can Build AI Research Applications Independently Using the API

Google launched the Gemini Deep Research research agent based on Gemini 3 Pro, opening its API for developers to embed SATA model research functions. It also integrates into multiple core services, demonstrating its practical value in professional fields.

AiBase Highlights:

🧠 Google launched Gemini Deep Research, offering APIs for developers to embed SATA model research functions for the first time.

📊 The new intelligent agent can handle large amounts of contextual data, applied in enterprise due diligence and drug toxicity research, among other areas.

🔥 Google integrates the deep research agent into core services like Google Search and Google Finance, driving the future vision of AI agents replacing humans in search tasks.

6、Cohere Launches Rerank 4: Four Times Larger Context Window, Enhancing Enterprise Search Accuracy

Cohere launched Rerank 4, which expands the context window to 32K, significantly improving enterprise search accuracy. The model comes in a fast version and a professional version, suitable for different scenarios. Additionally, Rerank 4 has self-learning capabilities, allowing users to customize the model based on common use cases without additional labeled data.

AiBase Highlights:

🌟 The context window expands to 32K, significantly improving retrieval accuracy.

🚀 Offers a fast version and a professional version to meet different scenario needs.

📈 First self-learning re-ranking model, allowing users to customize and improve performance.

7、Create 3D Figures Easily! Toplevel’s 3D Model Platform MakerWorld Integrates Tencent HuanYuan 3D 3.0 Large Model

The article introduces Toplevel's new 3D figure generator "PrintU," which relies on the Tencent HuanYuan 3D 3.0 large model to allow users to quickly generate high-quality, printable 3D models by uploading a single image. Traditionally, creating a 3D figure of a person required specialized equipment or a lot of manual modeling, but "PrintU" simplifies the process into three steps, greatly lowering the barrier to entry. Additionally, this technology introduces 3D-DiT hierarchical carving for the first time, improving modeling precision and detail, resulting in more realistic models.

AiBase Highlights:

🎨 Users just need to upload one image to quickly generate a personalized 3D figure.

🚀 The generator is based on Tencent HuanYuan 3D 3.0 technology, improving modeling precision and detail presentation.

🛠️ The creation process is simplified into three steps, lowering the barrier to 3D modeling.

More details: https://makerworld.com.cn/zh/makerlab/printU?from=makerlab

8、Google NotebookLM Upgrades Are Imminent: Ultra Members Get 50x Limit + No Watermark PPT, Accelerating the AI Note-Taking Revolution

Google's AI note-taking tool, NotebookLM, is about to receive a major update, with significant expansions in Ultra member privileges, indicating a clear trend toward catering to premium users in AI-assisted productivity tools. As an intelligent note generation platform, NotebookLM's upgrade not only enhances its generating capabilities but also optimizes user experience, helping professionals and creators efficiently handle complex information.

AiBase Highlights:

🧠 Generation limit increases by 50 times, eliminating bottlenecks.

🤖 Integrated with top-tier Gemini model, intelligence is upgraded.

📄 Notes sources expand to 600, increasing data inclusiveness.