The HappyHorse REST API exposes the same LaMa-powered object removal used in the editor: submit an image and mask, receive the cleaned result. Per-image credits, no subscription, about $0.0003 per call at Enterprise volume.
Catalog pipelines, listing platforms and DAM workflows need cleanup to happen inline — not in a browser tab. The API gives your backend the same 1-2 second inpainting the editor uses, so cleanup becomes just another step in your processing chain.
API access is available on paid accounts. Generate keys in your dashboard, or contact us for volume commitments and custom integration support. Batch pipelines are a common starting point — see batch processing for the editor-side equivalent.
The same LaMa (Large Mask Inpainting) model used in the editor — an open-source model by Samsung AI Center, specialized in large-area background reconstruction.
Per image: 1 credit per processed image. Credits come from the same packs as the editor ($0.50-$30.00) and never expire. No subscription, no monthly minimum.
New accounts get 100 free credits that work for both editor and API calls, so you can test the endpoint before purchasing credits.
JPG, PNG and WebP images up to 2048px, with a mask marking the areas to remove.
No subscription. No credit card required. Credits never expire.
Start Free