Skip to content
Date and source

Dates follow the news timeline in Beijing time and include both endpoints.

Oct 21Wed
Oct 20Tue
Oct 17Sat
  1. FastVideo 视频模型 · 模型发布60

    FastVideo releases FastH3-QAD-v2 video-audio model

    FastVideo releases FastH3-QAD-v2, an 8-step video and audio model trained with NVFP4 weights and quantization-aware distillation (QAD), runnable on a single Blackwell GPU (RTX 5090, RTX PRO 6000) or DGX Spark.

    Why it matters: The 8-step video and audio model lands on a single Blackwell GPU with NVFP4 quantized weights, and error data versus the BF16 teacher model is provided.

Oct 15Thu
  1. Black Forest Labs 官方博客82

    Black Forest Labs releases FLUX 3 Image Klein open-weight model

    Black Forest Labs releases FLUX 3 Image Klein, a 5.6B-parameter open-weight model that supports text-to-image, image editing, and multi-reference generation with up to 10 reference images, and features native 4K generation, pixel-level editing, and bounding-box control.

    Why it matters: An Open-Source Image Model with 5.6B Parameters Brings 4K Generation and Pixel-Level Editing Down to 8GB VRAM, Allowing Evaluation of Local Image Generation and Editing Solutions.

  2. Black Forest Labs · FLUX 模型62

    Black Forest Labs Releases ComfyUI Repackaged Model Files for Flux 3 Image Klein

    Black Forest Labs has released ComfyUI repackaged model files for Flux 3 Image Klein on Hugging Face. The original model repository is black-forest-labs/flux-3-image-klein.

    Why it matters: The official documentation provides the file directory and multiple quantized versions of Flux 3 Image Klein in ComfyUI, making it easy for local deployers to choose as needed.

  3. Black Forest Labs · FLUX 模型62

    Black Forest Labs open-sources FLUX 3 Image Klein Prompt Upsampler

    Black Forest Labs releases FLUX 3 Image Klein Prompt Upsampler, a 9B vision-language model that expands brief image requests into detailed prompts better suited to FLUX 3 Image Klein.

    Why it matters: The official release open-sources a 9B prompt expansion model that can expand brief requests into complete prompts including lighting, composition, and subject details, and supports reference images and bounding boxes.

  4. Fish Audio · 语音模型54

    Fish Audio releases Drama 3 int4 quantized version, reducing size to 22.1 GB and publishing WER evaluations

    An unofficial int4 quantized version, drama-3-int4, has appeared for Fish Audio's 35B (3B active) multi-speaker TTS model Drama 3, with a download size of 22.1 GB, lower than the original 41.5 GB. The expert layers, attention, and deltanet layers are quantized to int4, while the shared expert MLP remains in bf16.

Oct 13Tue
  1. Black Forest Labs · FLUX 模型78

    Black Forest Labs releases FLUX 3 Image Klein open-weight model

    Black Forest Labs Releases FLUX 3 Image Klein, a 5.6B Open-Weight Model That Simultaneously Supports Text-to-Image, Single-Reference Image Editing, and Multi-Reference Image Combination Generation, with Qwen3-VL-4B-Instruct as the text encoder.

    Why it matters: The 5.6B open-weight model unifies text-to-image, single-image editing, and multi-image combination into a single model, and provides quantized versions with VRAM usage details, making it easier for self-deployers to assess implementation costs.

  2. Black Forest Labs · FLUX 模型76

    Black Forest Labs Open-Sources FLUX 3 Image Klein Base, 5.6B Weights Support Text-to-Image and Multi-Image Editing

    Black Forest Labs has released FLUX 3 Image Klein Base on Hugging Face, an open-weight model with 5.6B parameters. Text-to-image, single-reference image editing, and multi-reference image composition share the same model, with the text encoder being Qwen3-VL-4B-Instruct.

    Why it matters: The 5.6B open weights cover text-to-image, single-image editing, and multi-image composition simultaneously, and provide three tiers of VRAM usage—BF16, FP8, and NVFP4—making it easy to evaluate self-hosting costs.

Oct 12Mon
  1. 腾讯 ARC · 创作模型60

    Tencent ARC Open-Sources Multi-Person Video Generation Model Multics

    Tencent ARC released Multics on Hugging Face, achieving scalable multi-person video generation through layered communication. The repository includes the complete 4-player Multics DiT checkpoint, configuration, Wan2.2 VAE, UMT5 text encoder, and tokenizer; the same checkpoint supports 6-, 8-, and 10-person generation without retraining. The weights are distributed under the Apache 2.0 license.

    Why it matters: Tencent ARC open-sources the multi-person video generation model Multics, where the same weights can scale to 10 people, letting creators evaluate generation options for multi-person scenes.

Oct 9Fri
  1. 通义千问 · 图像与音频模型62

    Alibaba releases Qwen-Image-2.1-Turbo accelerated image generation model

    Qwen released Qwen-Image-2.1-Turbo, an accelerated checkpoint of Qwen-Image-2.1 for text-to-image and image editing, requiring only 8 denoising steps.

    Why it matters: 8-step sampling and KV cache reuse of text and reference image context let creators assess the speedup potential of their image generation and editing workflows.

  2. Hugging Face Blog46

    Building six small models on Hugging Face with ML Intern: from a 0.8B prompt rewriter to a camera-angle LoRA

    Because the official Qwen-Image 2.1 prompt rewriter is a 9B model requiring about 20 GB of memory, the author used ML Intern in HuggingChat to build a 0.8B version that runs on CPU, achieves a 99.7% valid output rate, and consumes about one quarter of the teacher model's tokens, with a total compute cost of $16 including 8,797 annotations.

Oct 3Sat
  1. FastVideo 视频模型 · 模型发布62

    FastVideo releases the FastH3 Trim 8-step audio-video generation model

    FastVideo released the experimental model FastH3 Trim, which removes 8 of the least impactful blocks from FastH3 V2's 50 MiniMax H3 transformer blocks and replaces each block's AdaLN timestep projection with a shared rank-16 basis, trained with 8-step DMD2 (checkpoint 300), enabling 8-step text-to-video generation with synchronized audio.

    Why it matters: FastH3 Trim uses pruning plus shared AdaLN to bring 8-step audio-video generation to consumer-grade hardware, and can be compared against V2 to weigh image quality against size.

Sep 30Wed
Sep 29Tue
  1. Kyutai · 音视频模型20

    Kyutai Releases OVIE-ft-re10k: A Novel View Synthesis Fine-Tuned Model for RealEstate10K Scenes

    Kyutai has released OVIE-ft-re10k on Hugging Face. This is a checkpoint of OVIE fine-tuned for 50K steps on RealEstate10K multi-view data, accounting for only 2% of its 2-million-step pretraining budget. The architecture is identical to the base model, and it can be loaded directly by reusing the same code path.

  2. Lightricks · LTX 模型71

    Lightricks Releases LTX-2.5-22B IC-LoRA Alpha Gen Matting Model

    Lightricks released IC-LoRA Alpha Gen based on LTX-2.5-22B, which takes ordinary RGB video as input and generates frame-aligned grayscale alpha mattes without green screens, masks, or prompts, and can handle semi-transparent elements such as hair, fur, smoke, fire, sheer fabric, glass, and water.

    Why it matters: It generates frame-aligned alpha mattes from ordinary video without green screens or prompts, offering a new alternative for matting and compositing workflows.

Sep 28Mon
  1. Lightricks · LTX 模型62

    Lightricks releases the LTX-2.5-22b IC-LoRA SDR-to-HDR model

    Lightricks releases LTX-2.5-22b-IC-LoRA-SDR-To-HDR, which converts 8-bit SDR video to 16-bit HDR and outputs EXR frames in addition to an HLG master.

    Why it matters: Lightricks turns SDR-to-HDR into an IC-LoRA based on LTX-2.5 and provides a ComfyUI workflow and EXR output path, so it can be integrated directly into post-production pipelines.

Sep 27Sun
  1. Lightricks · LTX 模型71

    Lightricks releases the LTX-2.5-22b IC-LoRA old-film restoration model

    Lightricks releases LTX-2.5-22b-IC-LoRA-Restore on Hugging Face, an In-Context LoRA based on a reference video that renders old footage — low-resolution, low-bitrate, tape, or black-and-white/sepia scans — into clean, color, high-definition footage, removing compression damage, flicker, dirt, and scratches.

    Why it matters: Lightricks open-sources an old-film restoration LoRA, providing a two-stage workflow from 480-line scans to 4K and measured gains, which can be compared with existing restoration pipelines.

Sep 21Mon
  1. 腾讯 ARC · 创作模型60

    Tencent ARC Releases WorldCrafter-Fast Video Generation Weights

    Tencent ARC released self-contained WorldCrafter-Fast weights on Hugging Face, supporting image-to-video, text-to-video, and an interactive image-to-video demo, with inference possible without WorldCrafter-Base files.

    Why it matters: WorldCrafter-Fast has released standalone weights, with sampling step configurations for both image-to-video and text-to-video made public, making it easier for creators to evaluate integration costs.

Sep 17Thu
Sep 10Thu
Sep 3Thu
Aug 25Tue
  1. Hugging Face Blog62

    Gradio launches gr.Workflow: turning AI pipelines into a drag-and-drop canvas with auto-generated APIs

    Gradio's built-in gr.Workflow describes AI pipelines as graphs of typed nodes, offering a drag-and-drop canvas where each node can run and intermediate results are visible; the same graph is also a REST API and can be deployed to Hugging Face Spaces with one click.

    Why it matters: Gradio turns multi-step AI pipelines into a visual graph and automatically exposes REST interfaces, giving teams a basis to assess migration costs against their existing production workflows.

Aug 11Tue
Jul 23Thu
  1. Hugging Face Blog62

    Diffusers natively supports Nunchaku 4-bit diffusion inference

    A Hugging Face blog post explains that Diffusers now natively supports Nunchaku's 4-bit diffusion inference; loading a quantized checkpoint only requires calling from_pretrained(), with no local CUDA compilation needed.

    Why it matters: Nunchaku's 4-bit quantized inference is now natively integrated into Diffusers, allowing creators to evaluate the trade-off between VRAM usage and speed for image generation on consumer-grade GPUs.

Jul 6Mon
  1. Hugging Face Blog22

    PRX Part 4: Our Data Strategy

    The fourth installment of the PRX series reveals its data pipeline: pre-training data is composed of a mix of public and internal datasets, with VLMs used to regenerate long image descriptions, which are then converted into corpora suitable for streaming training. The team switched to the Qwen3-VL text encoder and computes text latents in real time during training, with measured throughput loss of approximately 3–4% (roughly one extra day for a 30-day training run). Images are uniformly encoded at JPEG quality 92, and initial re-encoding is nearly imperceptible in practice.

Jul 1Wed
Jun 9Tue
Apr 28Tue
  1. Hugging Face Blog36

    NVIDIA Releases Nemotron 3 Nano Omni: A Long-Context Multimodal Model for Document, Audio, and Video Agents

    NVIDIA has launched the Nemotron 3 Nano Omni omnimodal understanding model, featuring a Nemotron 3 hybrid Mamba-Transformer MoE backbone paired with the C-RADIOv4-H visual encoder and the Parakeet-TDT-0.6B-v2 audio encoder, supporting 100+ page documents, long audio and video understanding, and agentic computer operation.

Apr 9Thu
  1. Hugging Face Blog62

    Overworld Releases Waypoint-1.5 Real-Time Video World Model

    Overworld releases the real-time video world model Waypoint-1.5, which can generate real-time environments at up to 720p and 60 FPS on desktop hardware such as RTX 3090 to 5090, and adds a 360p tier for broader consumer-grade hardware, with Apple Silicon Mac support to follow.

    Why it matters: Waypoint-1.5 brings real-time world models down to consumer-grade GPUs, with 720p and 360p tiers letting creators evaluate the feasibility of local interactive generation.

Feb 3Tue
  1. Hugging Face Blog22

    PRX Text-to-Image Model Training Design: Lessons from Ablation Studies

    The Hugging Face blog publishes the second installment of the PRX text-to-image model training series, using ablation studies to document which training techniques actually work. The baseline is PRX-1.2B with 1.2B parameters, using pure Flow Matching, Flux VAE latent space, 256×256 resolution, a global batch size of 256, and 100k training steps, with data consisting of 1 million synthetic images generated by MidJourneyV6.

Jan 29Thu
  1. Hugging Face Blog40

    Gradio Team Launches Daggr: Chaining AI Workflows with Python Code and Auto-Generating Visual Canvases

    The Gradio team open-sources Daggr, a Python library that lets you chain Gradio apps, ML models, and custom functions with code and auto-generate visual canvases. It supports inspecting any node's output, re-running individual steps, configuring fallback nodes, and automatically saving workflow state and cached results. Daggr requires Python 3.10 or later, can be installed via pip, and workflows can be deployed to Hugging Face Spaces for sharing.

Jan 27Tue
  1. Hugging Face Blog22

    Architecture Choices in China's Open-Source AI Ecosystem: How MoE, Multimodality, and Domestic Hardware Can Surpass DeepSeek

    Chinese open-source models have nearly unanimously shifted to MoE architecture, with Kimi K2, MiniMax M2, and Qwen3 all adopting this approach, aiming for sustainable deployment at controlled costs. Open-source activity expanded to multimodality and agents starting February 2025, with StepFun's Step-Audio-R1.1 voice model outperforming closed-source models, and Tencent open-sourcing Hunyuan Video and Hunyuan 3D.

Jan 20Tue
  1. Hugging Face Blog65

    Overworld Releases Real-Time Interactive Video Diffusion Model Waypoint-1

    Overworld releases the real-time interactive video diffusion model Waypoint-1, controllable via text, mouse, and keyboard. Weights are now available on Hugging Face Hub, including Waypoint-1-Small, with Waypoint-1-Medium coming soon.

    Why it matters: The real-time interactive video diffusion model is open-sourced, allowing evaluation of its real-world performance in interactive world generation on consumer-grade hardware.

Jan 6Tue