Technical Specifications of bria/eraser
| Specification | Details |
|---|---|
| Model ID | bria/eraser |
| Provider | BRIA |
| Category | Image editing / object removal |
| Primary function | Removes selected objects or regions from an image and fills the erased area to match surrounding content |
| Input type | Image plus mask |
| Masking options | User-drawn mask or masks generated with BRIA’s object mask generator |
| Output | Edited image returned at the original resolution |
| Preservation behavior | Areas outside the provided mask remain unchanged; alpha can be preserved when supported by the endpoint configuration |
| Related BRIA capability | Part of BRIA’s Image Editing API suite under content manipulation tools |
| Typical use cases | Product photo cleanup, unwanted object removal, creative retouching, automated editing workflows |
| API style | HTTPS image editing endpoint in BRIA’s API platform |
What is bria/eraser?
bria/eraser is BRIA’s object-removal image editing model, exposed through BRIA’s image editing API as the Eraser capability. It is designed to remove unwanted elements or specific masked regions from an image, then regenerate the cleared area so the result blends naturally with the surrounding visual context. BRIA documents it as part of its specialized content-manipulation tooling for image editing workflows.
Unlike general prompt-based editing, the Eraser workflow is mask-driven: you specify exactly what should be removed, either by drawing over the image or by using BRIA’s mask generation tools to identify removable objects automatically. BRIA also notes that the edited image is returned at the original resolution and that regions outside the mask are preserved, which makes the model suitable for production use cases that require predictable, localized edits.
In practice, bria/eraser fits teams that need scalable, commercially oriented image cleanup for e-commerce, media operations, creative tooling, or automated asset preparation. BRIA positions the Eraser capability alongside SDKs, APIs, and embedded integrations for high-volume editing workflows.
Main features of bria/eraser
- Mask-based object removal: The model removes only the region specified by the provided mask, giving developers precise control over which object or area is erased.
- User-guided or automatic masking: Masks can come from direct user interaction, such as brush strokes in an editor, or from BRIA’s
/objects/mask_generatorworkflow for automated object selection. - Original-resolution output: BRIA states that the edited image is returned at the same resolution as the input, avoiding automatic downscaling during the erase operation.
- Localized edit preservation: Areas outside the selected mask remain unchanged, which is useful for deterministic, production-safe image editing pipelines.
- Transparency handling: BRIA indicates that when the input includes an alpha channel and the relevant setting is enabled, original transparency can be preserved in the output.
- Workflow-friendly integration: BRIA positions Eraser as part of a broader image editing API suite, making it suitable for integration with adjacent capabilities like mask generation, background operations, and generative fill.
- Commercial workflow focus: BRIA markets the Eraser capability for scalable editing, operational efficiency, and commercially ready deployment in products and business workflows.
How to access and integrate bria/eraser
Step 1: Sign Up for API Key
To access bria/eraser, first create an account on CometAPI and generate your API key from the dashboard. After signing in, store the key securely and use it in the Authorization header for all requests to the bria/eraser API.
Step 2: Send Requests to bria/eraser API
Use CometAPI's Bria-compatible endpoint at POST /bria/image/edit/eraser.
curl https://api.cometapi.com/bria/image/edit/eraser \
-H "Content-Type: application/json" \
-H "Authorization: Bearer $COMETAPI_API_KEY" \
-d '{
"prompt": "A paper boat floating on calm water at sunrise.",
"num_results": 1
}'
Step 3: Retrieve and Verify Results
The API returns a task ID. Poll GET /bria/{task_id} to check processing status and retrieve the output image URL when complete.