ComfyUI API Nodes Wave 2 is LIVE!

ComfyUI.org

2025-05-29 14:54:27

We’re excited to announce that ComfyUI API Nodes Wave 2 is now live — and it’s a big one!

Whether you're generating 3D assets, crafting text prompts with AI, or compositing media across open-source and API models, this update makes it all seamless — right inside ComfyUI.

Models Supported

LLMs

Google Cloud | Gemini 2.5 Pro, Gemini 2.5 Flash
OpenAI | GPT-4.1 / 4o / o1 / o3 / o4

Image / Video

RunwayML | Gen-4 Image and Image Reference, Gen-4 Turbo, Gen-3 Alpha Turbo
Vidu | Video Generation (coming soon)

Hyper3D AI | Hyper3D.AI - Rodin
Tripo AI | Tripo 3D V1.4 / 2.0 / 2.5

LLM API Nodes Unlocked

LLM nodes are here. Write, reason, prompt, and control pipelines — all from inside your ComfyUI workflow.

Use language models to dynamically generate text, build logic branches, adjust prompts, or even scaffold multi-step chains of actions across other nodes.

Gemini: Video Input and Description Example Workflow

Upgraded 3D Support

We’ve brought true 3D asset generation into your graph, with nodes that support both geometry and texture workflows — no more bouncing between tools.

Hyper3D AI | Rodin

High quality mesh fidelity with detailed presets and polygon count controls. Choose from: Sketch, Regular, Smooth, or Detailed — and output low-poly or high-res meshes.

Example: Rodin Multiview-to-3D workflow

Tripo AI

Highly configurable 3D creation pipeline. Supports text/image/multiview input with tools for texture editing, rigging, mesh refinement, and format conversion — all in-node.

Tripo Image-to-3D + Flux .1 Example Workflow

Mix & Match Online APIs with Open Models

Combine API models with open-source tools in powerful, custom pipelines. Whether you're prototyping or pushing to production, it’s all just node logic now.

RunwayML

Advanced showcase: Runway Image-to-image + OpenAI Chat + Flux dev hires fix + Runway Image-to-Video