Image Enhancement API: upscaling and background removal

Compare image super-resolution and background-removal workflows through Muapi’s unified image API. This repository groups two closely related developer tasks—upscale an image and isolate its subject—so a product, catalog, or creative pipeline can use one set of HTTP conventions.

Open the Image Upscaler API page · Open the Background Remover API page · Get an API key

Related Projects

• Awesome AI Image Models — image-model discovery and comparisons.
• Open Generative AI — curated open generative-AI tools.
• LoRA Trainer API examples — related image-model customization workflows.
• Muapi Python — general Python API integration.
• Muapi CLI — command-line access to Muapi models.

What this repository covers

• General-purpose, restoration-focused, and premium image upscaling.
• Image and video background removal as separate endpoint families.
• curl and Python examples for asynchronous image jobs and result polling.
• Input URL, resolution/upscale-factor, audio-preservation, and output-storage notes.

Model comparison

Task	Endpoint	Best for	Main input
Image upscale	ai-image-upscaler	General image enlargement	image_url plus model options
Image upscale	seedvr2-image-upscale	Restoration and detail recovery	image_url, optional resolution
Image upscale	topaz-image-upscale	Premium image enhancement	image_url, optional upscale_factor
Image background removal	ai-background-remover	Fast subject isolation	image_url
Video background removal	video-background-remover	Frame-accurate video matting	video_url, background/output options

Use the Image Upscaler page for super-resolution selection and the Background Remover page for image/video matting details and current pricing.

Quick start with curl

Upscale an image:

export MUAPI_API_KEY="your_key"
export IMAGE_URL="https://example.com/product-small.jpg"

curl -X POST "https://api.muapi.ai/api/v1/ai-image-upscaler" \
  -H "x-api-key: $MUAPI_API_KEY" \
  -H "Content-Type: application/json" \
  -d "{\"image_url\":\"$IMAGE_URL\"}"

Remove an image background:

curl -X POST "https://api.muapi.ai/api/v1/ai-background-remover" \
  -H "x-api-key: $MUAPI_API_KEY" \
  -H "Content-Type: application/json" \
  -d "{\"image_url\":\"$IMAGE_URL\"}"

Poll either submission using its request identifier:

export REQUEST_ID="id_from_the_submission_response"
curl "https://api.muapi.ai/api/v1/predictions/$REQUEST_ID/result" \
  -H "x-api-key: $MUAPI_API_KEY"

See examples/quickstart.py for a complete image-upscale polling loop.

Choosing an enhancement workflow

• Use general upscaling for routine catalog and social-media enlargement.
• Use SeedVR2 when restoring detail in a degraded or low-resolution source is the priority.
• Use Topaz when a premium enhancement tier or explicit scale factor is more important than lowest cost.
• Use image background removal for still assets and video-background-remover for moving footage; do not send a video URL to an image endpoint.
• Store the returned file URL in durable storage if it is needed after the job completes.

Troubleshooting

Problem	Check
Output is soft or over-sharpened	Choose the model for the source quality and avoid repeatedly upscaling the same file.
Transparent background is unexpected	Inspect the result’s format/alpha behavior before compositing.
Video request fails	Use the video endpoint and verify duration, codec, and output-container options.
Result URL disappears	Download or copy it to durable storage immediately after success.

License

Documentation and examples are MIT licensed. Input assets, generated outputs, and model-provider usage remain subject to applicable terms.