Will It Run AI · Workflow VRAM Calculator

Build your AI workflow and check if it fits in VRAM.

Stack a base model with ControlNets, LoRAs, and IP-Adapter — see the total VRAM breakdown in real time. No more guessing whether Flux + ControlNet + 2 LoRAs will fit on your GPU.

Build Your Workflow

HardwareBase Model (required)

Precision

Resolution

Workflow Components

ControlNet

+1.2 GB each · Depth, Canny, Pose, etc.

LoRA

+164 MB each · Style, character, concept

IP-Adapter

+0.5 GB · Image prompt conditioning

VRAM Breakdown

Runs natively

Your workflow needs 8 GB / 24 GB available

Required: 8.0 GBAvailable: 24.0 GB

Component Breakdown

Base Model (Stable Diffusion XL 1.0)2.6B params @ FP16

8.0 GB

Total8.0 GB

Stable Diffusion XL 1.0

2.6B parameters · UNET · 30 steps

Available ControlNets: Canny Edge, Depth Map, OpenPose, IP-Adapter, Union (Multi-Control)

LoRA ecosystem: large · 5,000 on CivitAI

Common Workflows