Skip to main content

​
Overview

Generate a short reel-sized film from just content descriptions.
  • Output Type: video
  • Base Model: flux-dev
  • Estimated Cost: 500 credits
  • Handler: modal

​
Parameters

​
Required Parameters

​
prompt
string
required
Describe what your reel is about
  • Label: Prompt

​
Optional Parameters

​
agent
string
Optional username of the agent who is writing the reel.
  • Label: Agents
​
use_voiceover
boolean
default:true
Include a voiceover in the reel
  • Label: Voiceover
​
voice
string
Override voice. If left blank, voice is auto-generated.
  • Label: Voice
​
voiceover
string
Override voiceover. If left blank, voiceover text is auto-generated.
  • Label: Voiceover text
​
use_music
boolean
default:true
Include music in the reel
  • Label: Music
​
music_prompt
string
Override music prompt. If left blank, music prompt is auto-generated.
  • Label: Music prompt
​
min_duration
integer
default:10
Minimum number of seconds for the reel
  • Label: Minimum duration
  • Minimum: 5
  • Maximum: 60
​
orientation
string
default:"portrait"
The aspect ratio of the reel
  • Label: Orientation
  • Options: landscape, portrait
​
use_lora
boolean
default:false
Apply LoRA finetune model style to image generation
Models created with Eden LoRA trainer can add people, styles and conceptual embeddings into the diffusion model, giving it an idea of new information provided by the user.
  • Label: Use LoRA
​
video_model
string
default:"high"
The quality of the video model to use. Low = Fast and cheap, but lowest quality. High = Best quality, but slow and expensive.
Default to highest quality unless otherwise specified. Runway = low, Kling = medium, Veo2 = high
  • Label: Video model quality
  • Options: low, medium, high
​
lora
lora
Use a LoRA finetune on top of the base model.
Only visible when use_lora=true
  • Label: LoRA
​
lora_strength
float
default:0.6
Strength of the LoRA
If outputs resemble the LoRA but have low prompt adherence or all look the same, turn down the LoRA strength.
Only visible when use_lora=true
  • Label: LoRA Strength
  • Minimum: 0.0
  • Maximum: 1.5