AI News
Daily curated AI news: breakthroughs, product launches, industry trends and more.
Latest News
Total 174 articles
AI News (2026/2/11): Domestic AI Interactive Application Loopit Receives Public Praise...
The domestic AI application Loopit has sparked heated discussions on overseas social media, attracting numerous users with its unique "AI interactive content generation" feature. Developed by the team led by Weipeng Chen, co-founder of BaiChuan Intelligence, Loopit allows users to generate clickable, scrollable, and interactive content through text descriptions, and supports secondary creation of others' works. Recently, Loopit received public praise from Tesla CEO Elon Musk.
AI News (2026/2/10): ByteDance Launches Seedream 5.0 Image Generation Model
ByteDance has officially launched the Seedream 5.0 image generation model, which supports 2K/4K resolution output, improves the accuracy of prompt understanding, detail texture rendering, and intelligent reasoning capabilities, and adds an online search function. The model is now available for a limited-time free trial on platforms such as Jianying, CapCut, and Xiaoyunque.
AI News (2026/2/9): ByteDance Launches New AI Video Generation Model Seedance 2.0
ByteDance has recently released its new AI video generation model, Seedance 2.0. This model supports multimodal reference generation, capable of handling up to 12 reference files, including images, videos, and audio. The core breakthrough of Seedance 2.0 lies in the integration of AI generation and post-editing functions, allowing users to directly modify unsatisfactory parts, thereby significantly reducing the rate of unusable content.
AI News (2026/2/6): OpenAI Releases New Programming Model: GPT-5.3-Codex
On February 6, OpenAI released the new programming model GPT-5.3-Codex, which achieved State-of-the-Art (SOTA) results in the SWE-Bench Pro and Terminal-Bench 2.0 tests, with a programming score 11.9% higher than Claude Opus 4.6. GPT-5.3-Codex has the ability to debug, deploy, and operate office software, with a 25% speed improvement and the capability to participate in its own development optimization.
AI News (2026/2/5): Keling AI Officially Enters the 3.0 Era!
Keling AI has officially upgraded to version 3.0, introducing a new series of Keling 3.0 models. This upgrade integrates multi-modal input and output, redefining the AI video workflow. The Video 3.0 model has been deeply optimized for narrative and control, adding an intelligent storyboarding feature that can automatically schedule shot sizes and camera positions. Image 3.0 Omni introduces new deep narrative and batch creation features for image sets.
AI News (2026/2/4): Qwen3-Coder-Next Open-Sourced: Small and Powerful!
The Tongyi large model has announced the open-sourcing of the Qwen3-Coder-Next programming AI model. This model uses an 80B parameter MoE (Mixture of Experts) architecture, with only 3B parameters activated during each inference, significantly reducing computational requirements. On the SWE-Bench Verified benchmark, Qwen3-Coder-Next achieved a problem-solving rate of over 70%, performing on par with larger dense models.
AI News (2026/2/3): GLM-OCR Released: SOTA Performance for Complex Documents
Zhipu AI has officially released and open-sourced the professional-level OCR model GLM-OCR, achieving "small size, high precision" document parsing capabilities with just 0.9B parameters. The model topped the authoritative OmniDocBench V1.5 evaluation with a score of 94.6, demonstrating state-of-the-art (SOTA) performance in tasks such as text recognition, formula recognition, table parsing, and information extraction. It is particularly optimized for challenging scenarios like handwritten text, complex tables, and code documents, supporting batch processing of PDFs and images with a throughput of 1.86 pages per second.
AI News (2026/2/2): Step 3.5 Flash Launches! Faster, Stronger, and More Stable Agent...
Step 3.5 Flash, the new open-source base model from Jieyue Xingchen, is designed specifically for Agent scenarios. The model uses a sparse MoE architecture with 196 billion parameters, activating only about 11 billion parameters per token, and supports a context length of 256K. Key advantages include a maximum inference speed of 350 TPS, performance comparable to closed-source models in Agent tasks and mathematical reasoning, and significant efficiency improvements through MTP-3 technology.
AI News (2026/1/31): Open-Source UnifoLM-VLA-0 Large Model to Enhance General Humanoid...
Yushu Technology has announced the open-source release of the visual-language-action large model UnifoLM-VLA-0. Based on the Qwen2.5-VL-7B architecture, the model has been trained with 340 hours of real-world data, integrating 2D/3D spatial perception and dynamics prediction capabilities to overcome the limitations of traditional VLMs in physical interactions.
AI News (2026/1/30): SenseNova-MARS Open-Sourced by SenseTime: Breaking the Ceiling of...
SenseTime has open-sourced the multimodal autonomous reasoning model SenseNova-MARS, offering 8B and 32B versions. The model outperforms Gemini-3-Pro (69.06 points) and GPT-5.2 (67.64 points) with a comprehensive score of 69.74 points in core benchmarks such as MMSearch and HR-MMSearch, becoming the first open-source Agentic VLM to support dynamic visual reasoning and deep integration of image and text search.
Subscribe to Daily AI News
Get curated AI news delivered to your inbox daily. Never miss an update.
We never share your email. Unsubscribe anytime.