Contexts Optical Compression

🚀 「大模型」1小时从0训练26M参数的视觉多模态VLM！🌏 Train a 26M-parameter VLM from scratch in just 1 hours!

🚀🚀 「大模型」2小时完全从0训练26M的小参数GPT！🌏 Train a 26M-parameter GPT from scratch in just 2h!

SeC: Advancing Complex Video Object Segmentation via Progressive Concept Construction

Comfyui implementation of OpenIXCLab Sec-4B

Provide with pre-build flash-attention package wheels on Linux and Windows platforms using GitHub Actions

Kyutai's Speech-To-Text and Text-To-Speech models based on the Delayed Streams Modeling framework.

Lynx: Towards High-Fidelity Personalized Video Generation

Open-source framework for building AI-powered apps in JavaScript, Go, and Python, built and used in production by Google

Text-audio foundation model from Boson AI

A fast AI Video Generator for the GPU Poor. Supports Wan 2.1/2.2, Qwen Image, Hunyuan Video, LTX Video and Flux.

SoTA open-source TTS

A unified inference and post-training framework for accelerated video generation.

[NeurIPS 2025] Training-Free Efficient Video Generation via Dynamic Token Carving

[NeurIPS 2025] Let Them Talk: Audio-Driven Multi-Person Conversational Video Generation

Taming Stable Diffusion for Lip Sync!

Connect any AI model to 600+ integrations; powered by MCP 📡 🚀

DFloat11: Lossless LLM Compression for Efficient GPU Inference

Open-source unified multimodal model

[CVPR 2025 Highlight] UltraFusion: Ultra High Dynamic Imaging using Exposure Fusion

[CVPR 2025 Highlight] Real-time dense scene reconstruction with SLAM3R

[NeurIPS 2025] Image editing is worth a single LoRA! 0.1% training data for fantastic image editing! Surpasses GPT-4o in ID persistence~ MoE ckpt released! Only 4GB VRAM is enough to run!

SANA: Efficient High-Resolution Image Synthesis with Linear Diffusion Transformer

ComfyUI Plugin of Nunchaku

Model Compression Toolbox for Large Language Models and Diffusion Models

[ICLR2025 Spotlight] SVDQuant: Absorbing Outliers by Low-Rank Components for 4-Bit Diffusion Models

ACE-Step: A Step Towards Music Generation Foundation Model

WhisperPlus: Faster, Smarter, and More Capable 🚀

Expanding FramePack into a multifunction video creation tool