ComfyUI Releases

A minor patch release that adds support for the Kling V3-Turbo video generation model.

Jun 18, 2026

This release focuses on 3D visualization with new Gaussian Splat and Point Cloud nodes, alongside significant enhancements to the Asset Management API and updates for various partner video nodes.

Jun 16, 2026

The release brings Ideogram V4 model support, updates to Depth Anything 3, and fixes for Triposplat and DINO3 models, alongside improvements to PiD and math expression nodes.

Jun 3, 2026

This release introduces significant performance improvements through multi-threaded loading and GPU acceleration, alongside new AI model support like PixelDiT, PiD, and Tripo3D, along with new 3D and utility nodes.

Jun 1, 2026

Focuses on expanding audio and video capabilities with new model integrations like MoGe and Stable Audio 3, alongside improvements to the LTX video generation workflow.

May 20, 2026

A maintenance release adding new partner nodes for Flux, Grok, and OpenAI, alongside support for the HiDream-O1-Image model and extended Save3D capabilities.

May 13, 2026

A massive feature release introducing text generation for Gemma4, new video models like Cogvideox and Void, and significant workflow improvements like auto-regressive video generation and causal window handling.

May 11, 2026

Adds support for image upscaling models (SUPIR), frame interpolation (RIFE/FILM), and the latest AI video models (Veo 3 Lite, SAM 3.1), alongside bug fixes for Veo resolution handling.

Apr 27, 2026

A minor maintenance release focusing on updates to partner nodes, specifically adding new SVG models and making certain outputs optional in 3D text-to-3D nodes.

Apr 17, 2026

Addresses a critical memory issue in quantized models and introduces a new utility node for JSON extraction, along with options to disable default templates.

Apr 17, 2026

A small update fixing issues with the Ernie model and adding support for the Ministral text model and Sonilo audio nodes.

Apr 16, 2026

This release introduces support for several new AI models including LTX2, Qwen3.5, Ace Step 1.5 XL, and RT-DETRv4, alongside new utility nodes like Curve and Number Convert. It also includes backend optimizations for fp8 support and RAM cache integration.

Apr 13, 2026

A minor maintenance release with no detailed changelog provided in the text.

Mar 25, 2026

Focuses on critical bug fixes for fp16 stability, including the Canny node, sampling processes, and WAN VAE rendering issues.

Mar 23, 2026

A major update adding FP16 intermediate support and mxfp8 precision. It also improves VRAM efficiency for LTX and WAN VAEs and adds support for AMD gfx1150 (Strix Point).

Mar 21, 2026

A minor maintenance release with no detailed changelog provided.

Mar 15, 2026

A minor maintenance release with no detailed changelog provided.

Mar 13, 2026

Introduces a modular asset scanner architecture and supports Flux 2 Klein KV cache models. Includes the reversion of the external distributed caching API.

Mar 13, 2026

Adds a new Math Expression node and updates API nodes to support TencentSmartTopology and Gemini 3.1 Flash-Lite models. Fixes fp16 audio encoder issues.

Mar 7, 2026

Fixes for LTX2 vocoder and LTX audio VAE to improve handling in low VRAM scenarios.

Mar 5, 2026