Create

The Create menu provides AI-powered tools for generating and processing content. Tools connect to cloud GPU services (RunPod, WaveSpeed, Eleven Labs) or run locally.

Create Module

The Create module (create) manages the AI content generation panel.

Module Parameters

Parameter	Type	Description
`visible`	bool	Create panel visibility

Content Types

Type	Description
Image	AI image generation (Flux, SDXL)
Video	AI video generation (WAN, Hunyuan, LTX)
Audio	Text-to-speech (RunPod Qwen TTS)
Audio (ElevenLabs)	High-quality TTS via Eleven Labs API
Video (WaveSpeed)	Cloud video generation via WaveSpeed API

Commands

Command ID	Name	Description
`create:toggle-visible`	Show Create	Toggle the Create panel
`create:show-image`	Create Image	Open Create panel in Image mode
`create:show-video`	Create Video	Open Create panel in Video mode
`create:show-audio`	Create Audio (TTS)	Open Create panel in TTS mode (RunPod)
`create:show-audio-elevenlabs`	Create Audio (Eleven Labs)	Open Create panel in Eleven Labs mode
`create:show-wavespeed`	Create Video (WaveSpeed)	Open Create panel in WaveSpeed mode

Lua API

-- Toggle create panel
pl.cmd("create:toggle-visible")

-- Open specific creation modes
pl.cmd("create:show-image")
pl.cmd("create:show-video")
pl.cmd("create:show-audio")
pl.cmd("create:show-audio-elevenlabs")
pl.cmd("create:show-wavespeed")

Image Generation

Generate images from text prompts using diffusion models hosted on RunPod (via ComfyUI) or the built-in image generation panel.

Models

Model	ID	Description
Flux Schnell	`flux-schnell`	Fast generation (Flux)
Flux Dev	`flux-dev`	Higher quality (Flux)
SDXL Base	`sdxl-base`	Stable Diffusion XL base
SDXL Refiner	`sdxl-refiner`	SDXL with refinement pass

Image Prompt Parameters

Parameter	Type	Default	Description
`prompt`	string	—	Text description of desired image
`negative_prompt`	string?	—	What to avoid in the image
`style`	string?	—	Style hint (“photo”, “digital painting”, “anime”)
`seed`	u64?	random	Seed for reproducibility
`model`	Model	Flux Schnell	Generation model
`orientation`	Orientation	Square	Image dimensions
`num_steps`	u32	25	Inference steps (higher = more detail)
`guidance_scale`	f32	7.5	Prompt adherence (CFG scale)
`num_outputs`	u32	1	Number of images to generate

Orientation Options

Option	Aspect
Square	1:1
Landscape	16:9
Portrait	9:16

Text-to-Speech

RunPod TTS

Generate speech from text using Qwen TTS voices hosted on a RunPod container. Generated audio imports directly into your project as an asset.

Requires: Running TTS container

TTS Commands

Command ID	Description
`tts:toggle-visible`	Toggle TTS Manager panel
`tts:import-audio`	Import generated audio to timeline
`tts:register-asset`	Register TTS output as a project asset
`tts:save-download`	Save TTS audio to disk

-- Open TTS manager
pl.cmd("tts:toggle-visible")

-- Import generated audio to timeline
pl.cmd("tts:import-audio", {
  url = "https://...",
  name = "Narration Take 1"
})

Eleven Labs TTS

High-quality AI voice generation with wider voice selection and voice cloning. Requires an Eleven Labs API key in Settings.

Video Generation

RunPod Video (WAN / Hunyuan / LTX)

Generate video clips from text or images using AI models on RunPod.

Command ID	Description
`videogen:toggle-visible`	Toggle Video Generator panel
`videogen:import-video`	Import generated video to timeline
`videogen:save-download`	Save video to disk
`videogen:load-history`	Load generation history
`videogen:save-history`	Save generation history

-- Open video generator
pl.cmd("videogen:toggle-visible")

-- Import generated video
pl.cmd("videogen:import-video", {
  url = "https://...",
  name = "AI Scene"
})

Requires: Running Video Gen container

WaveSpeed

Cloud-based video generation — no GPU container needed. See WaveSpeed.

Tools

Tool	Description	Requires
Split Audio	AI stem separation (vocals, drums, bass, other) via Demucs	TTS container
Transcribe	Speech-to-text via Whisper	TTS container
Audio to MIDI	Pitch detection → MIDI notes	Local
Face Animation	Facial animation from video/Live Link Face	Local
Image Edit	Built-in paint tool	Local
3D Editor	Three.js-based scene editor	Local
Shader Browser	Browse and apply ISF shaders	Local

Split Audio

Uses the same TTS container (which includes Demucs) to separate audio into stems:

Stem	Description
Vocals	Lead and background vocals
Drums	Percussion and drums
Bass	Bass instruments
Other	Everything else

Infrastructure

Service	Description	Details
Container Manager	RunPod GPU pod lifecycle	Containers
ComfyUI	Node-based AI workflows	ComfyUI
WaveSpeed	Cloud video generation	WaveSpeed
LoRA Manager	Fine-tuning model weights	Requires ComfyUI pod

Keyboard Shortcuts

Shortcut	Action
`G`	Open Workflow Editor (Generate)

Containers — RunPod GPU container management
ComfyUI — Node-based AI workflows
WaveSpeed — Cloud video generation