Academia

This page is auto-generated from model configurations. Last updated: 2026-07-01.

This reference lists all available Academia video generation models and their parameters. Use these parameter names when calling the Generation API.

• Foley Control
• MM Audio
• MM Audio 2
• Reverse Video
• Video Foreground Extractor - BiRefNet v2

---

Foley Control

Foley Control is a model that automatically generates synchronized sound effects for videos, using text prompts to shape the type of sound while matching the timing and action on screen.

Model ID: model_controlfoley

Capabilities: video2video

LLM Markdown: https://app.scenario.com/api/models/model_controlfoley/markdown

Parameter	Type	Required	Default	Min	Max	Allowed Values	Description
video	file	Yes	-	-	-	-	Video to generate synchronized audio for.
prompt	string	No	-	-	-	-	Optional text prompt describing the desired audio. When combined with the video it provides text-controlled video-to-audio generation; leave empty for pure video-to-audio
negativePrompt	string	No	-	-	-	-	Audio characteristics to avoid.
referenceAudio	file	No	-	-	-	-	Optional 2-4 second reference audio clip whose timbre should guide the generated audio. Audio shorter than 2s is zero-padded; longer is truncated.
duration	number	No	8	1	30	-	Target audio duration in seconds. Truncated to the source video length when shorter.
numInferenceSteps	number	No	25	4	100	-	Number of steps.
guidanceScale	number	No	4.5	0	20	-	Classifier-free guidance strength.
maskAwayClip	boolean	No	false	-	-	-	Disable the CLIP visual stream for text-driven generation.
seed	number	No	-	0	2147483647	-	Optional seed for reproducibility.

MM Audio

Model ID: model_mm-audio

Capabilities: video2video

LLM Markdown: https://app.scenario.com/api/models/model_mm-audio/markdown

Parameter	Type	Required	Default	Min	Max	Allowed Values	Description
prompt	string	Yes	-	-	-	-	Text prompt for generated audio
video	file	Yes	-	-	-	-	Video file for video-to-audio generation
negativePrompt	string	No	-	-	-	-	Negative prompt to avoid certain sounds
duration	number	No	30	1	30	-	Output duration in seconds. If this value exceeds the video’s length, the video’s full duration will be used instead.
numSteps	number	No	25	4	50	-	The number of steps to generate the audio for
cfgStrength	number	No	4.5	1	20	-	Higher values will keep output closer to the prompt
seed	number	No	-	-1	-	-	Random seed. Use -1 or leave blank to randomize the seed

MM Audio 2

MMAudio generates synchronized audio given video and text prompts. It can be combined with video models to get videos with audio.

Model ID: model_mm-audio-2

Capabilities: video2video

LLM Markdown: https://app.scenario.com/api/models/model_mm-audio-2/markdown

Parameter	Type	Required	Default	Min	Max	Allowed Values	Description
prompt	string	Yes	-	-	-	-	Text prompt for generated audio
video	file	Yes	-	-	-	-	Video to generate the audio for.
negativePrompt	string	No	-	-	-	-	Negative prompt to avoid certain sounds
duration	number	No	8	1	30	-	Output duration in seconds.
numSteps	number	No	25	4	50	-	The number of steps to generate the audio for
cfgStrength	number	No	4.5	1	20	-	Higher values will keep output closer to the prompt
maskAwayClip	boolean	No	false	-	-	-	Mask away certain sounds in the audio
seed	number	No	-	0	65535	-	Random seed for reproducible generation

Reverse Video

Reverse a video so it plays backwards, including both video and audio streams.

Model ID: model_reverse-video

Capabilities: video2video

LLM Markdown: https://app.scenario.com/api/models/model_reverse-video/markdown

Parameter	Type	Required	Default	Min	Max	Allowed Values	Description
video	file	Yes	-	-	-	-	The video you want to play backwards. Both the picture and the sound are reversed. Accepts MP4, MOV, WebM, M4V, and GIF.

Video Foreground Extractor - BiRefNet v2

BiRefNet is a model that cleanly separates the main subject from the background in a video, producing clear and detailed cutouts.

Model ID: model_birefnet-v2-video

Capabilities: video2video

LLM Markdown: https://app.scenario.com/api/models/model_birefnet-v2-video/markdown

Parameter	Type	Required	Default	Min	Max	Allowed Values	Description
video	file	Yes	-	-	-	-	Video to remove background from
model	string	No	General Use (Light)	-	-	General Use (Light), General Use (Light 2K), General Use (Heavy), Matting, Portrait	Model to use for background removal. The ‘General Use (Light)’ model is the original model used in the BiRefNet repository. The ‘General Use (Light)’ model is the original model used in the BiRefNet repository but trained with 2K images. The ‘General Use (Heavy)’ model is a slower but more accurate model. The ‘Matting’ model is a model trained specifically for matting images. The ‘Portrait’ model is a model trained specifically for portrait images. The ‘General Use (Light)’ model is recommended for most use cases.
operatingResolution	string	No	1024x1024	-	-	1024x1024, 2048x2048	The resolution to operate on. The higher the resolution, the more accurate the output will be for high res input images.
outputMask	boolean	No	false	-	-	-	Whether to output the mask used to remove the background
refineForeground	boolean	No	true	-	-	-	Refine the foreground for better results
videoQuality	string	No	high	-	-	low, medium, high, maximum	Video output quality
videoWriteMode	string	No	balanced	-	-	fast, balanced, small	Video write mode for encoding
videoOutputType	string	No	VP9 (.webm)	-	-	X264 (.mp4), VP9 (.webm), PRORES4444 (.mov), GIF (.gif)	Output format. Transparency (alpha): VP9 (.webm) and ProRes 4444 (.mov). MP4 has no alpha. GIF: limited transparency only.