ComfyUI April 2026 Update Wrap-Up: 12 New Models, Parallel Job Execution, and ComfyHub Milestone

The ComfyUI team has released a comprehensive wrap-up of all platform updates shipped in April 2026, including 12 new native and partner models, parallel workflow execution support, and a major milestone for the ComfyHub workflow directory. This release expands creative capabilities for video, image, audio, vector, and restoration workflows, while improving efficiency for batch processing and production pipelines. The update offers new opportunities for workflow builders and template authors curating and sharing work on platforms like ComfyUITemplates.com.

What ComfyUI Is Announcing

Per the ComfyUI Blog post dated May 4, 2026, the team shipped a wide range of new tools and platform improvements throughout April 2026. The announcement covers 12 new models across multiple creative use cases, one major workflow feature update, and growth for the official ComfyHub workflow hub.

Key New Models in the April 2026 Release

The following 12 models were added to ComfyUI as native or partner nodes in April 2026:

Seedance 2.0: ByteDance’s next-gen video model, live as a partner node on day one of availability. Offers three launch modes: Text-to-Video, Reference-to-Video, and First/Last-Frame to Video, with adjustable creative control. Supports uploading AI-generated images or real human portraits to create videos with consistent identity and authentic facial expressions.
Happy Horse: Multi-talented partner node covering Text-to-Video, Image-to-Video, Reference-to-Video, and video editing.
Wan 2.7: Latest iteration of the Wan 2.7 multi-use case video model, available as a partner node.
Ernie-Image: Baidu’s open-source text-to-image model built on an 8B Diffusion Transformer (DiT), with fully open weights for user customization.
Ace Step 1.5 XL: Open-source music generation model with commercial-grade quality, tuned for full-song coherence including verses, choruses, and complete arrangements.
Quiver: Partner node for structured SVG generation, producing vector output rather than pixel-based renders. Ideal for icons, diagrams, and use cases requiring infinite scalability without re-rendering.
GPT Image 2: OpenAI’s latest image model, available as a partner node, supporting both text-to-image generation and image editing in a single workflow.
SUPIR: Open-source image restoration tool designed to repair blurry, degraded, or low-resolution images. Pairs well with standard upscaling workflows.
RIFE & FILM Frame Interpolation: Open-source frame interpolation models now with native ComfyUI support. Enables users to smooth motion, achieve higher frame rates, or extend short clips into cinematic-length content.
SAM 3 & SAM 3.1: Meta’s SAM 3 and SAM 3.1 models for image and video segmentation, now natively supported in ComfyUI.
Veo 3.1 Lite & Veo 3.1 Fast: Google’s lighter, lower-cost Veo model variants, now supported as partner nodes.
Sonilo: Video-to-music partner node that generates soundtracks tuned to the motion and mood of input video clips.

Key Feature Releases: Parallel Job Execution

The only new workflow feature announced in the April 2026 update is Parallel Job Execution:

Enables users to run multiple workflows in parallel simultaneously.
Delivers significant speedups for batch processing, testing workflow variations, and running production pipelines.
The ComfyUI team notes this feature is "the single feature most likely to change how your day actually feels" for users managing multiple jobs.

ComfyHub Growth Milestone

The official ComfyHub workflow directory reached a major milestone in April 2026:

Crossed nearly 500 total workflows, per the announcement.
The ComfyUI team notes that users may find pre-built workflows for their needs on ComfyHub, reducing duplicate work for common use cases.

Why This Matters for Workflow Creators and Template Authors

This update has direct implications for creators building and sharing ComfyUI workflows, including those curating templates for ComfyUITemplates.com:

Expanded model support means template authors can build and share new workflows for 12 additional models across video, image, audio, vector, and restoration use cases, increasing the diversity of available templates.
Parallel Job Execution streamlines iteration for template authors: testing multiple workflow variations simultaneously reduces development time for new templates.
ComfyHub’s growing library of nearly 500 workflows gives creators a resource to find existing templates, which they can also list on ComfyUITemplates.com to reach a wider audience.

How This Affects ComfyUI Templates and Apps

Template authors curating work on ComfyUITemplates.com can now add new workflows for the 12 models and Parallel Job Execution feature announced in the April 2026 update.
The growth of ComfyHub underscores demand for high-quality, pre-built workflows, making directories like ComfyUITemplates.com increasingly valuable for users seeking vetted templates.
Partner node models (including Seedance 2.0, Happy Horse, and others) use standardized node structures, making it easier for template authors to build compatible workflows that work across the ComfyUI ecosystem.

ComfyUITemplates.com

April Wrapped