How Cubeflow Works

From idea to finished asset in three simple steps.

01

Pick your models

Choose from every best-in-class AI model spanning every creative medium. Image generators like Imagen 4 and Flux 2 Pro. Video creators like Veo 3 and Runway Gen-3. Music composers like Suno. Voice synthesizers like ElevenLabs. And LLMs like GPT-5.2 and DeepSeek Chat. We add new models within days of their release.

Model Selection Preview
02

Connect the nodes

Drag models onto an infinite canvas. Connect outputs to inputs with a single click. Each node’s output automatically feeds into the next model. Build linear pipelines or complex branching workflows with parallel processing.

Node Connection Preview
03

Hit Run

Execute your entire pipeline with one click. Watch results stream in real-time as each node completes. Preview images, video, and audio right inside the canvas. If something needs tweaking, adjust a parameter and re-run just that node.

Pipeline Execution Preview

Supported Models

Every top AI model, and counting

Google Imagen 4
Google Veo 3
Runway Gen-3
Kling 3.0
Suno Music
ElevenLabs
DeepSeek Chat
GPT-5.2
Seedream V4
Flux 2 Pro
GPT-4o Image
Kling 2.6

Ready to build your first pipeline?

Start with the free plan. No credit card required.

Start Building — Free