1. ChatGPT Now Has Its Own App Store
OpenAI has officially launched a new in-app directory, similar to an app store, that allows developers to connect their applications directly to ChatGPT conversations. This innovative feature aims to enhance the functionality and versatility of ChatGPT by enabling users to interact with a wide range of third-party apps within the AI’s interface. OpenAI has also indicated plans to introduce monetization options for these integrated applications in the near future, creating new revenue streams for developers and further expanding the ChatGPT ecosystem.
Read full story
2. Google Launches Faster, Cheaper Gemini 3 Flash AI Model
Google has unveiled Gemini 3 Flash, its latest artificial intelligence model designed for enhanced speed and cost-effectiveness. This new model will serve as the core technology for the Gemini application and Google’s AI Search functionalities. The launch underscores Google’s commitment to advancing AI capabilities, offering users a more efficient and accessible AI experience. Gemini 3 Flash aims to provide faster response times and reduce operational costs, making advanced AI more pervasive in everyday digital interactions. This development positions Google competitively in the rapidly evolving AI landscape, directly addressing the demand for efficient and powerful AI solutions.
Read full story
3. Grok API Now Features Real-Time Voice Capabilities
xAI has officially launched real-time voice integration for Grok’s API, enabling developers to create more dynamic and interactive conversational artificial intelligence agents. This new feature allows for streaming speech, which means AI agents can now process and respond to voice commands instantly, replicating more natural human-like conversations. This advancement is expected to significantly enhance user experience for applications built leveraging Grok’s AI technology, offering seamless voice interactions.
Read full story
4. Meta’s New VLM Halves Trainable Parameters
Meta has introduced a groundbreaking new vision-language model (VLM) that significantly rethinks traditional VLM mechanisms. This innovative model predicts continuous embeddings, a novel approach that dramatically optimises its architecture. By adopting this method, Meta’s new VLM successfully reduces the number of trainable parameters by a remarkable 50%, making it significantly more efficient and potentially faster to train and deploy. This development marks a substantial advancement in the field of artificial intelligence, promising more powerful and resource-friendly models for various applications.
Read full story
5. FLUX.2 [max]: Advanced AI for Stunning Image Generation
A new AI model, FLUX.2 [max], has been introduced, promising significant advancements in image generation capabilities. This cutting-edge tool is designed to provide users with real-time web grounding, ensuring that generated images are contextually relevant and accurate. Furthermore, FLUX.2 [max] excels in producing cinematic visuals, allowing for high-quality, professional-grade imagery. A key feature is its ability to generate consistent characters across multiple images, which is crucial for narrative coherence and brand consistency in various applications. This model is expected to be a game-changer for digital content creators, designers, and film professionals looking for top-tier AI image solutions.
Read full story
6. AI Model Predicts Fruit Fly Development with High Accuracy
Researchers at MIT have developed a groundbreaking deep-learning model that can accurately predict how fruit fly cells behave during development. This model achieves an impressive 90% accuracy in forecasting the intricate processes involved in fruit fly formation. This significant advancement holds immense potential for future medical research, particularly in enhancing our ability to detect and understand various diseases at a cellular level. By understanding fundamental biological processes more deeply, scientists can pave the way for new diagnostic tools and therapeutic interventions.
Read full story
7. Gemini & NotebookLM: Smarter AI Conversations
Google’s Gemini application now seamlessly integrates with NotebookLM, offering users enhanced AI conversational capabilities. This new connection allows NotebookLM users to incorporate their organised sources and research materials directly into their Gemini AI interactions, leading to more informed and context-rich discussions. This feature aims to streamline knowledge management and AI-powered learning for a more efficient user experience.
Read full story
8. Exa’s AI Revolutionises People Search with Superior Results
Exa has developed an advanced method for finding individuals online, showcasing its effectiveness through indexing over a billion profiles. The AI company has also released a benchmark report, demonstrating its superior performance in people search capabilities compared to existing solutions. This innovation promises more accurate and efficient results for users seeking specific individuals.
Read full story
9. LongCat-Video-Avatar: Create Realistic Talking Videos from Audio
LongCat-Video-Avatar is an innovative technology that transforms audio input into natural-looking talking videos. This AI-powered tool generates avatars with realistic expressions and gestures, making the video content highly engaging and lifelike. Users can simply provide an audio track, and the system will animate a video avatar to speak and react in a human-like manner, ideal for presentations, digital marketing, and educational content.
Read full story
10. Microsoft’s TRELLIS.2: High-Res 3D Model Generation from Images
TRELLIS.2 is an innovative tool developed by Microsoft that transforms 2D images into detailed 3D models. This advanced system creates textured 3D assets at a high resolution, incorporating efficient compression techniques. It is designed to streamline the process of generating realistic digital models for various applications, from virtual reality to game development and architectural visualisation. The technology ensures that the output is not only visually rich but also optimised for performance and storage.
Read full story
11. Qwen Code: AI Dev Tool with 2,000 Daily Requests
Qwen Code is a powerful Command Line Interface (CLI) tool designed to streamline AI development workflows. This innovative solution offers users up to 2,000 daily requests, making it a valuable asset for developers and researchers. Access to its features is securely managed via OAuth, ensuring both convenience and data protection for all users.
Read full story
12. HY-WorldPlay: AI for Interactive 3D World Creation
HY-WorldPlay introduces a cutting-edge streaming video diffusion model engineered to generate interactive 3D virtual worlds. This innovative technology allows for real-time control over the generated environments, offering developers and content creators powerful tools for dynamic world-building. Developers are encouraged to explore its capabilities for creating immersive and responsive applications.
Read full story
