AIGC Latest News
Stay up to date with latest AIGC news and updates.
ByteDance Releases Seaweed-7B: A Cost-Effective Video Generation Foundation Model
Seaweed-7B achieves performance surpassing 14B parameter models with only 7 billion parameters, at just one-third the training cost of industry standards, bringing new possibilities to the video generation field
FloED: Open Source Efficient Video Inpainting with Optical Flow-Guided Diffusion
A new video inpainting framework FloED has released its code and weights, achieving higher video coherence and computational efficiency through optical flow guidance
PixelFlow: Generative Models Working Directly in Pixel Space
PixelFlow innovatively operates in raw pixel space, simplifying the image generation process without requiring pre-trained variational autoencoders, enabling end-to-end trainable models
VAST-AI Releases HoloPart: Generative 3D Part Amodal Segmentation Technology
HoloPart can decompose 3D models into complete, semantically meaningful parts, solving editing challenges in 3D content creation
VAST-AI and Tsinghua University Open Source UniRig: A Framework for Automatic Skeleton Rigging of All 3D Models
UniRig uses autoregressive models to generate high-quality skeleton structures and skinning weights for diverse 3D models, greatly simplifying the animation workflow
OmniSVG: Fudan University and StepFun Launch Unified Vector Graphics Generation Model
OmniSVG is a new unified multimodal SVG generation model capable of producing highly complex, editable vector graphics from various inputs including text, images, or character references
TTT-Video: Technology for Long Video Generation
Researchers develop TTT-Video model using Test-Time Training technology based on CogVideoX 5B, capable of generating coherent videos up to 63 seconds long
ByteDance Releases UNO: Extending Generation Capabilities from Less to More
ByteDance Creative Intelligence team releases UNO model, unlocking greater controllability through in-context generation, achieving high-quality image generation from single to multiple subjects
EasyControl: A New Framework for Efficient and Flexible Control of Diffusion Transformer
Tiamat AI team releases EasyControl framework, adding conditional control capabilities to DiT models, now supported in ComfyUI via the ComfyUI-easycontrol plugin
HiDream-I1 Open Source Release - Next Generation Image Generation Model
HiDream.ai releases HiDream-I1, a new open-source text-to-image model with 17B parameters that outperforms existing open-source models in multiple benchmarks, supporting high-quality image generation in various styles
Hi3DGen: A New Framework for High-Fidelity 3D Geometry Generation through Normal Bridging
Stable-X team introduces Hi3DGen, an innovative framework for generating high-fidelity 3D models from images, addressing the lack of geometric details in existing methods through normal bridging technology
VAST AI Research Open Sources TripoSF: Redefining New Heights in 3D Generation Technology
TripoSF, based on the innovative SparseFlex representation, supports 3D model generation at resolutions up to 1024³, capable of handling open surfaces and complex internal structures, significantly improving 3D asset quality
Kunlun Wanwei Open-Sources SkyReels-A2: Commercial-Grade Video Generation Framework
Kunlun Wanwei releases the world's first commercial-grade controllable video generation framework SkyReels-A2, enabling multi-element video generation through dual-branch architecture, bringing new possibilities for e-commerce, film production and more
Alibaba's Tongyi Lab Releases VACE: Video Creation and Editing Enters Unified Era
Alibaba Group's Tongyi Lab introduces VACE, the world's first unified framework for diverse video tasks, covering text-to-video generation, video editing, and complex task combinations
StarVector: A Multimodal Model for SVG Code Generation
The StarVector project implements automatic generation of SVG vector graphics code from images and text, providing new creative tools for designers and developers.
ByteDance Releases InfiniteYou: Flexible Photo Recrafting While Preserving User Identity
ByteDance introduces InfiniteYou (InfU), an innovative framework based on Diffusion Transformers that enables flexible photo recrafting while preserving user identity, addressing limitations in existing methods regarding identity similarity, text-image alignment, and generation quality
StdGEN: Semantic-Decomposed 3D Character Generation from Single Images
Tsinghua University and Tencent AI Lab jointly introduce StdGEN, an innovative pipeline that generates high-quality semantically-decomposed 3D characters from single images, enabling separation of body, clothing, and hair
Stability AI Releases Stable Virtual Camera: Technology to Transform 2D Photos into 3D Videos
Stability AI launches new AI model Stable Virtual Camera, capable of converting ordinary photos into 3D videos with authentic depth and perspective effects, providing creators with intuitive camera control
Tencent Releases Hunyuan3D 2.0 - Innovative 3D Asset Generation System
Tencent launches Hunyuan3D 2.0 system with a two-stage process for generating high-quality 3D models, featuring multiple open-source model series that support creating high-resolution 3D assets from text and images
Kuaishou Launches ReCamMaster Monocular Video Reframing Technology
Kuaishou Technology has launched ReCamMaster, a generative video technology that allows users to create new camera perspectives and motion paths from a single video.
Open-Sora 2.0 Released: Commercial-Grade Video Generation at Low Cost
LuChen Technology releases Open-Sora 2.0 open-source video generation model, achieving performance close to top commercial models with just $200,000 in training costs
Ali Tongyi Lab Releases VACE: All-in-One Video Creation and Editing Model
Ali Tongyi Lab launches multifunctional video creation and editing model VACE, integrating various video processing tasks into a single framework to lower the barrier of video creation
Microsoft Releases ART Multi-layer Transparent Image Generation Technology
Microsoft Research introduces intelligent layered generation based on global text prompts, supporting creation of transparent images with 50+ independent layers
Tencent Open-Sources HunyuanVideo-I2V Image-to-Video Model
Tencent's Hunyuan team releases open-source model for generating 5-second videos from single images, featuring smart motion generation and custom effects
Alibaba Open-Sources ViDoRAG Intelligent Document Analysis Tool
Alibaba introduces a document analysis system capable of understanding both text and images, improving processing efficiency for complex documents by over 10%
THUDM Open-Sources CogView4 - Native Chinese-Supported DiT Text-to-Image Model
THUDM releases CogView4 open-source image generation model with native Chinese support, leading in multiple benchmark tests
Sesame Unveils CSM Voice Model for Natural Conversations
Sesame Research introduces dual-Transformer conversational voice model CSM, achieving human-like interaction with open-source core architecture
Alibaba's Wan2.1 Video Generation Model Officially Open-Sourced
Alibaba has officially open-sourced its latest video generation model, Wan2.1, which can run with only 8GB of video memory, supporting high-definition video generation, dynamic subtitles, and multi-language dubbing, surpassing models like Sora with a total score of 86.22% on the VBench leaderboard
Alibaba Releases ComfyUI Copilot: AI-Driven Intelligent Workflow Assistant
Alibaba International Digital Commerce Group (AIDC-AI) releases the ComfyUI Copilot plugin, simplifying the user experience of ComfyUI through natural language interaction and AI-driven functionality, supporting Chinese interaction, and offering intelligent node recommendations and other features
Alibaba's WanX 2.1 Video Generation Model to be Open-Sourced
Alibaba has announced that its latest video generation model, WanX 2.1, will be open-sourced in the second quarter of 2025, supporting high-definition video generation, dynamic subtitles, and multi-language dubbing, ranking first on the VBench leaderboard with a total score of 84.7%
Google Releases PaliGemma 2 Mix: An Open-Source Visual Language Model Supporting Multiple Tasks
Google introduces the new PaliGemma 2 mix model, supporting various visual tasks including image description, OCR, object detection, and providing 3B, 10B, and 28B scale versions
Skywork Opensources SkyReels-V1: A Video Generation Model Focused on AI Short Drama Creation
Skywork has open-sourced its latest video generation model, SkyReels-V1, which supports text-to-video and image-to-video generation, featuring cinematic lighting effects and natural motion representation, and is now available for commercial use.
Light-A-Video - A Video Relighting Technology Without Training
Researchers have proposed a new video relighting method, Light-A-Video, which achieves temporally smooth video relighting effects through Consistent Light Attention (CLA) and Progressive Light Fusion (PLF).
StepFun releases Step-Video-T2V: A 300 Billion Parameter Text-to-Video Model
StepFun has released the open-source text-to-video model Step-Video-T2V, which has 300 billion parameters, supports the generation of high-quality videos up to 204 frames, and provides an online experience platform
Kuaishou Introduces CineMaster: Breakthrough in 3D-Aware Video Generation
Kuaishou officially releases CineMaster text-to-video generation framework, enabling high-quality video content creation through 3D-aware technology
Alibaba Open Sources InspireMusic: An Innovative Framework for Music, Song and Audio Generation
Alibaba's latest open-source project InspireMusic, a unified audio generation framework based on FunAudioLLM, supporting music creation, song generation and various audio synthesis tasks.
Alibaba Open Sources ACE++: Zero-Training Character-Consistent Image Generation
Alibaba Research Institute open sources image generation tool ACE++, supporting character-consistent image generation from single input through context-aware content filling technology, offering online experience and three specialized models.
ByteDance Releases OmniHuman: Next-Generation Human Animation Framework
ByteDance research team releases OmniHuman-1 human animation framework, capable of generating high-quality human video animations from a single image and motion signals.
DeepSeek Open-Sources Janus-Pro-7B: Multimodal AI Model
Tencent Releases Hunyuan3D 2.0: Open-Source High-Quality 3D Generation Model and End-to-End Creation Engine
Tencent releases Hunyuan3D 2.0, open-sourcing the complete DiT model and launching a one-stop 3D creation engine with innovative features including skeletal animation and sketch-to-3D, revolutionizing metaverse and game content creation
ComfyUI Project Two-Year Anniversary
ComfyUI celebrates its second birthday, evolving from a personal project into the world's most popular generative AI visual tool. Happy birthday, ComfyUI!
NVIDIA Open Sources Sana - An AI Model for Efficient 4K Image Generation
NVIDIA releases new Sana model capable of quickly generating images up to 4K resolution on consumer laptop GPUs, with ComfyUI integration support
ByteDance Open Sources LatentSync - High-Precision Lip Sync Technology Based on Diffusion Model
ByteDance releases open-source lip sync tool LatentSync, based on audio-conditioned latent space diffusion model, enabling precise lip synchronization for both real people and animated characters while solving frame jittering issues common in traditional methods
VMix: ByteDance Introduces Innovative Aesthetic Enhancement Technology for Text-to-Image Diffusion Models
ByteDance and University of Science and Technology of China jointly launch VMix adapter, enhancing AI-generated image aesthetics through cross-attention mixing control technology, seamlessly integrating with existing models without retraining
Tencent Open Sources StereoCrafter: One-Click 2D to 3D Video Conversion
StereoCrafter, jointly developed by Tencent AI Lab and ARC Lab, is now open source. It can convert any 2D video into high-quality stereoscopic 3D video, supporting various 3D display devices including Apple Vision Pro
LuminaBrush: AI Lighting Editor Released by ControlNet Creator
lllyasviel, creator of ControlNet and IC-Light, releases LuminaBrush - a new AI tool that enables precise lighting control through a two-stage framework with intuitive brush interactions
Genesis: Breakthrough Universal Physics Engine and Generative AI Platform Released
Genesis project releases a new universal physics engine and generative AI platform, integrating physics simulation, robot control and generative AI capabilities to provide comprehensive solutions for robotics and physical AI applications
Odyssey Releases Explorer - Breakthrough Generative World Model
Odyssey launches Explorer, the first generative world model that can convert 2D images into complete 3D worlds, supporting dynamic effects and mainstream creative tool editing, bringing revolutionary changes to film, gaming and other fields
Mining Virus Found in ComfyUI Impact-Pack Plugin - Urgent Action Required
Popular ComfyUI plugin Impact-Pack discovered to contain mining malware through Ultralytics package, affecting numerous users. This article details the virus situation and solutions.
Luma Launches Photon - A Revolutionary AI Image Generation Model with Exceptional Value
Luma introduces revolutionary Photon and Photon Flash image generation models, redefining AI creation with outstanding value and superior image quality
Tencent Open Sources HunyuanVideo - A New Era in Video Generation
Tencent officially open sources HunyuanVideo, the industry's largest video generation model with 13 billion parameters, achieving leading performance in video quality and motion stability
Stability AI Releases Stable Diffusion 3.5 Large ControlNet Models
Stability AI launches three new ControlNet models for Stable Diffusion 3.5 Large, including Blur, Canny, and Depth, providing more precise control capabilities for image generation
NVIDIA Releases Edify 3D - A Revolutionary 3D Asset Generation Technology
NVIDIA has launched the new Edify 3D technology, capable of generating high-quality 3D assets in just two minutes, including detailed geometry, clear topology, high-resolution textures, and PBR materials.
Lightricks Releases Real-Time Video Generation Model LTX-Video
Lightricks launches the LTX-Video real-time video generation model based on DiT, supporting real-time generation of high-quality videos, and has been open-sourced on GitHub and Hugging Face.
InstantX Releases FLUX.1-dev IP-Adapter Model
InstantX team has released the IP-Adapter model based on FLUX.1-dev, bringing more powerful image reference capabilities to the FLUX model
FLUX Official Tools Suite Released
Black Forest Labs releases a variety of FLUX official tools, including local redraw, ControlNet, and image style conversion
IC-Light V2 Release: Enhanced Image Editing Capabilities
IC-Light V2 is a Flux-based image editing model that supports various stylized image processing, including oil paintings and anime styles. This article details the new features and applications of IC-Light V2.
Stable Diffusion 3.5 Launches: A New Era in AI Image Generation
Stability AI releases Stable Diffusion 3.5, offering multiple powerful model variants, supporting commercial use, and leading the market in image quality and prompt adherence.
ComfyUI V1 Released: Cross-Platform Desktop App with One-Click Installation
ComfyUI has released Version 1, offering a cross-platform desktop application with one-click installation, a revamped user interface, and numerous feature improvements, significantly enhancing the user experience.
Kuaishou and PKU Jointly Release Pyramidal Flow Matching Video Generation Model
Kuaishou Technology and Peking University jointly developed the Pyramidal Flow Matching model, an autoregressive video generation technology based on flow matching, capable of producing high-quality, long-duration video content.
Jasperai Releases Flux.1-dev ControlNet Model Series
Jasperai introduces a series of ControlNet models for Flux.1-dev, including surface normals, depth maps, and super-resolution models, providing more precise control for AI image generation.
ComfyUI-PuLID-Flux: Implementing PuLID-Flux in ComfyUI
ComfyUI-PuLID-Flux is an open-source project that integrates PuLID-Flux into ComfyUI, offering powerful image generation and editing capabilities.
Meta Introduces Movie Gen: AI Models for Video and Audio Generation
Meta introduces the Movie Gen series of models, including video generation, audio generation, and personalized video editing capabilities, opening new frontiers in AI content creation.
ostris Releases OpenFLUX.1: A Commercially Usable De-distilled Version of FLUX.1-schnell
ostris has released OpenFLUX.1, a de-distilled version of the FLUX.1-schnell model that can be fine-tuned, uses normal CFG values, and retains commercial licensing.
Black Forest Lab Launches FLUX 1.1 [Pro]
Black Forest Labs introduces FLUX1.1 [pro], a generative AI model featuring ultra-fast generation, superior image quality, and 2K ultra-high resolution generation, opening unprecedented opportunities for creators, developers, and businesses.
THUDM Open Sources New Image Generation Models: CogView3 and CogView-3Plus
THUDM has open-sourced their latest image generation models CogView3 and CogView-3Plus-3B, showcasing exceptional performance and efficiency. CogView3 utilizes cascaded diffusion technology, while CogView-3Plus-3B is a lightweight model based on the DiT architecture, bringing significant breakthroughs to the field of text-to-image generation.