LoRA Pilot Logo
🚀 Just add GPU and go

Civitai-Like Simplicity
Full Pro Stack Included

LoRA Pilot makes training and inference approachable for beginners with clean wizards, while still exposing the full firepower of AI Toolkit, Kohya, ComfyUI, InvokeAI, Diffusion Pipe, and TensorBoard.

Deploy on RunPod View Source
Supports SD 1.5/2.1 • SDXL • SD3 • FLUX.1 (dev/schnell/kontext) • Wan • OmniGen2 • PixArt • AuraFlow • LTX/LTX2 • Z-Image-Turbo • and more
Release notes live on GitHub and inside the product.

Why LoRA Pilot?

Because stable diffusion tooling should feel like a product, not a scavenger hunt.

Training + Inference In One Place

AI Toolkit, Kohya, ComfyUI, InvokeAI, Diffusion Pipe, TensorBoard, Jupyter, and more. No setup marathon before first result.

Simple Wizards, Full Controls

Beginner-friendly flows replace chaotic GUIs, while power users keep access to the tools they already trust.

Built For Real Workloads

Persistent workspace paths, reliable startup behavior, stable logs/telemetry, and service controls that do not randomly forget your choices.

What's in the box?

A Complete AI Studio

ComfyUI

Your node-based playground. ComfyUI Manager pre-installed for easy custom nodes.

Kohya SS

The industry standard for training LoRAs, complete with a dedicated Web UI.

AI Toolkit UI

First-class trainer with UI on port 8675. Start/stop jobs through Supervisor, no duct tape scripts.

VS Code (Web)

Full IDE in your browser. Edit code, scripts, and configs without leaving the container.

JupyterLab

For when you need to get nerdy with data science notebooks.

TagPilot

Dataset tagger embedded right into the workflow. Organize your training data fast.

InvokeAI

A powerful, polished interface for creative professionals. The "Diva" of the bunch.

GitHub Copilot Chat

Global chat drawer in ControlPilot with prompt execution, status updates, and output right in the UI.

Diffusion Pipe

Structured training pipelines with better control over experiments and reproducible runs.

TensorBoard

Track loss curves, training metrics, and runs without going full spreadsheet detective.

See It In Action

Real workflows, real controls, no terminal gymnastics required.

Main Dashboard
ControlPilot Dashboard

ControlPilot Dashboard

Run your full stack from one control plane.

Models Manager
Model Downloads

Models Manager

Download and organize models without CLI juggling.

TagPilot
TagPilot Interface

TagPilot

Clean tags faster so training data stops fighting back.

Workflow Tools
ComfyUI Integration

Workflow Tools

ComfyUI, InvokeAI, Kohya, and more in one workspace.

AI Toolkit
AI Toolkit Training

AI Toolkit Training

Configure and launch trainer jobs with a first-class flow.

Copilot Chat
ControlPilot Copilot Chat

Copilot In-App

Run prompts in UI with status and output where you work.

Powering Training & Inference For

Stable Diffusion 1.5 Stable Diffusion 2.1 SDXL SD3 FLUX.1-dev FLUX.1-schnell FLUX.1-kontext OmniGen2 PixArt Alpha PixArt Sigma AuraFlow Lumina2 HunyuanVideo LTX LTX2 Wan 2.1 / 2.2 Cosmos Qwen-Image Qwen-Image-Edit Z-Image-Turbo

Designed for User Experience

We believe training and inference should be as easy as Civitai. Whether you prefer clean wizards or CLI depth, LoRA Pilot covers both.

CLI Power Users

  • • pilot status to check services
  • • models pull sdxl-base to get models
  • • Custom scripts for workflow automation

GUI Lovers

  • • ControlPilot central dashboard
  • • One-click model downloads
  • • Visual training configuration with TrainPilot

Frequently Asked Questions

Short answers for the stuff everyone asks first.

What is LoRA Pilot?

LoRA Pilot is an all-in-one containerized workspace for training and inference with both wizard-style GUI flows and CLI workflows.

Which tools are already included?

AI Toolkit, Kohya SS, ComfyUI, InvokeAI, Diffusion Pipe, TensorBoard, ControlPilot, JupyterLab, TagPilot, plus service orchestration and persistent workspace paths.

Do I need Linux or CLI expertise?

No. Wizard-based flows are designed for users with little or zero Stable Diffusion experience, and advanced controls are there when you want them.

Where can I find release notes and changelog?

On GitHub and inside the product.