ComfyUI Releases

ComfyUI v0.43.0 is a broad feature release adding expanded MOSS-image and FIBO 2.0 model support, many new Partner Nodes, and core workflow improvements. It also includes performance work for Qwen 3.x, dynamic VRAM KV caches, AMD VAE attention, and Flux2 VAE handling. New utility nodes and cloud/API-related additions round out the release.

Oct 27, 2026

This release addresses stability issues where specific users encountered crashes while utilizing the MOSS image template workflow. The update aims to improve the reliability of the software for those working with complex image generation setups.

Oct 19, 2026

This release introduces support for new AI models like MOSS-image and Krea 3, alongside performance optimizations for Qwen 3.x. It also includes updates to workflow templates, 3D model remeshing improvements, and fixes for the webcam and text generation nodes.

Oct 18, 2026

This release introduces new nodes for HDR processing and latent geometry extraction, alongside support for the Flux3-image model and MiniMax-H3 LightVAE. It also includes updates to workflow templates and a core route for the default workflow.

Oct 15, 2026

v0.40.0 introduces a default-on assets system with metadata extraction and EXR preview support, alongside expanded 3D model handling and camera workflow tools. The release also features significant performance optimizations for Qwen 3.5 LLMs, new partner node integrations, and improvements to the Linear model API.

Oct 13, 2026

v0.39.0 addresses critical stability issues regarding database locking during asset scans and saves. It introduces new command-line options for offline usage and partner node management, along with support for new video and image generation partners like Ideogram and ElevenLabs.

Oct 5, 2026

This release focuses on performance optimizations for models like Qwen and Flux, alongside significant improvements to asset management and database locking to prevent UI freezes. It also adds support for several new partner model nodes and features like HDR color space conversion.

Sep 29, 2026

This release focuses on performance improvements and new model support, introducing MoGe 3, Qwen-image 2.1, and Meshy 7.1 support. It also enhances Qwen model performance through CUDA graphs and fixes several GPU-related stability issues.

Sep 21, 2026

This release adds core features like Generic Loops and Marigold v2, alongside support for new AI models such as Yue2 and BFL Flux Video Edit. It also enhances AMD support, reorganizes model blueprints, and introduces new partner node integrations.

Sep 15, 2026

ComfyUI v0.35.0 introduces the Comfy Compiler for performance optimization and expands support for new partner models like WAN3-Prime and Claude Fable 5.1. Key additions include robust 3D workflow tools, advanced video editing capabilities, and critical fixes for memory management and alpha channel handling.

Sep 9, 2026

This release focuses on performance improvements and new model support, including HDR video capabilities and several partner integrations like Wan 3.0 and Pixverse V6, alongside optimizations for dynamic VRAM and Gemma4.

Aug 26, 2026

This release focuses on fixing stability issues with nested latents and KSamplerAdvanced, alongside updates to workflow templates and partner nodes for MiniMax and Bria.

Aug 13, 2026

A major update raising the minimum supported PyTorch version to 2.7 and introducing support for new video and image models like LTX 2.5, Qwen-Image 3.0 Pro, and Grok-Imagine-Image-2.0.

Aug 11, 2026

This release enhances performance and adds support for Flux 3 video, Wan-Animate2, and various partner models including TopazAI and ByteDance's SeeDance 2.5.

Aug 8, 2026

Introduces MiniMax-H3 support and support for the faster PrunaVAED decoder for LTX2.3, alongside new partner nodes for Ideogram, Minimax, and xAI.

Aug 3, 2026

A maintenance release focusing on frontend fixes and the addition of new API and partner nodes without major architectural changes.

Jul 31, 2026

Focuses on security fixes, adding Seed Audio 1.0 and Seedream 5 Pro support, and introducing basic text overlay and 3D advanced nodes.

Jul 29, 2026

A significant update addressing security vulnerabilities and performance, while dropping support for older PyTorch versions and adding advanced nodes for text and 3D workflows.

Jul 15, 2026

This release focuses on performance optimization with comprehensive support for Int8 models on Turing GPUs, alongside new integrations from partner nodes including Alibaba, Grok, ByteDance, and Google.

Jun 30, 2026

The update introduces support for Qwen3-VL text generation and new video generation capabilities with the Kling V3-Turbo and Luma Rays 3.2 models, while improving 3D asset loading and node organization.

Jun 23, 2026