Search AI Products and News

Lightricks releases open-source LTX-Video13B, a 13B-parameter video generation model with multi-scale rendering, achieving 30x faster speeds. It runs on consumer GPUs, supports 1216×704 real-time generation, and offers text/image/video-to-video modes. The model enhances coherence and detail, enabling keyframe control and style transfer. Free for SMEs, it includes training tools and optimized versions to democratize AI video creation.....

Moonshot AI's open-source model Kimi K2 excels in programming, matching Qwen3-235B-A22B and nearing o3-mini-high & Claude-3.7-Sonnet. With 1T MoE, 128k context, 65.8% accuracy on SWE-bench, and $0.14/M tokens, it's ideal for coding. Supports web generation, workflows, vLLM/Hugging Face deployment, MIT-licensed.....

NVIDIA's Canary-Qwen-2.5B sets a 5.63% WER record on Hugging Face OpenASR. This CC-BY licensed model combines FastConformer encoder with Qwen3-1.7B LLM decoder for efficient speech-to-text and NLP. Supports multi-GPU deployment for cloud/edge applications.....

MirageLSD, the world's first AI real-time video conversion model with 40ms latency, enables instant scene/outfit changes via gestures. Applied in gaming/livestreaming, its LSD model uses Diffusion Forcing to eliminate long-generation errors.....

Google launched Veo3 API for text-to-video & audio generation, but costs are high ($0.75/s for 720p). Used in gaming cinematics & 3D modeling. Faster 'Veo3Fast' mode not yet API-accessible. Requires Gemini API & Google Cloud subscription.....

Baidu AI Assistant introduces 'Video Call' feature, enabling real-time video chats via Baidu APP. It covers four scenarios: object/insect identification, styling advice, pet behavior analysis, and emotional support, with dialect recognition for elderly users. This upgrade transforms AI from a tool to a smart life companion.....

Recently, the ByteDance Seed team collaborated with the University of Hong Kong and Fudan University to introduce an innovative reinforcement learning training method called POLARIS. This method successfully enhances the mathematical reasoning capabilities of small models to levels comparable to those of large models through a carefully designed Scaling RL strategy, offering a new approach for optimizing small models in the field of artificial intelligence. Experimental results show that the 4 billion parameter open-source model Qwen3-4B trained using POLARIS achieved remarkable performance on AIME25 and AIME24 mathematical tests.

On July 11, PixVerse AI video creation platform, which has surpassed 60 million global users, announced a major feature upgrade — the addition of the 'Multi-Keyframe Generation' function in the Start-End Frame module. This marks a new stage in AI video creation, transitioning from the generation of single segments to narrative expression. Users can now upload up to 7 images as keyframes via the web version's start-end frame feature, and the AI will automatically analyze the semantic relationships between frames, intelligently building smooth action and scene transition paths. This technological breakthrough enables static images to be presented dynamically.

Meta and Berkeley developed StreamDiT for real-time AI video generation: 1) 16fps 512p on single GPU, 4B-param model creates 1-min videos with live edits; 2) Novel buffer enables parallel processing (2 frames/0.5s) with 8-step optimization; 3) Trained on 3K HD videos + 2.6M dataset; 4) Outperforms rivals in motion smoothness, 30B-param version shows quality potential; 5) Enables interactive video despite transition flaws.....

Recently, Tencent Cloud officially launched the API service for the Tencent Hunyuan A13B model on its official website. The input price is set at 0.5 yuan per million Tokens, and the output price is 2 yuan per million Tokens, which has quickly sparked enthusiastic discussions in the developer community. As the first 13B-level MoE (Mixture of Experts) open-source hybrid inference model in the industry, Hunyuan-A13B features a total of 80B parameters and only 13B activated parameters, achieving performance comparable to leading open-source models of the same architecture, while also demonstrating efficient reasoning capabilities.

Shanghai Jiao Tong University and DP Technology's X-Master AI system set a 32.1 record in the HLE test, featuring tool-enhanced reasoning and multi-agent collaboration for biomedical breakthroughs.....

A major breakthrough has emerged in the AI video field — Vidu Q1 video model, launched by Shengshu Technology, officially released the Reference Generation feature, offering a revolutionary experience where video material can be generated from imagination in one step, redefining the technical boundaries and production efficiency of content creation. In traditional video production processes, creators have to go through complex steps such as script writing, character design, storyboard drawing, real scene shooting, and post-production editing, and the creation of a short film often takes several weeks or even months. The release of the Reference Generation feature in Vidu Q1 has completely broken this established model. Users

PolyU & OPPO developed DLoRAL, a video super-resolution framework using dual LoRA: CLoRA for temporal consistency and DLoRA for spatial details. Its two-stage training balances quality and speed (10× faster inference), with open-source code/models available. Limited in tiny text recovery but promising for real-time applications.....

Hong Kong Polytechnic University and OPPO Research Institute jointly released the open-source video super-resolution framework DLoRAL, which generates high-definition videos in one step using diffusion models. The framework adopts a dual LoRA architecture: C-LoRA maintains temporal consistency between frames, while D-LoRA enhances spatial details. Through a two-stage training strategy, it optimizes temporal coherence and high-frequency information. Compared to traditional methods, DLoRAL improves inference speed by 10 times while maintaining smoothness, significantly enhancing image details, and providing an efficient open-source solution for video HD enhancement.

Tencent open-sourced 'Hunyuan-A13B', an 80B-parameter MoE model with dynamic inference (130B active params). Supports 256K context, trained on 20T tokens. Achieves 87.3% on AIME2024 math, but comparisons show version disparities.....