Skip to content
Last updated

Bria Products Overview

Image Generation

Generate high-quality, commercially-safe visuals using Bria's state-of-the-art models.

Next-Generation Image Generation (Recommended):

Our latest generation API offers precise control via structured JSON prompts.

  • How it Works: A VLM Bridge translates your text/image input into a detailed JSON structured_prompt. An advanced image model then uses this prompt for deterministic image generation.
  • VLM Options:
    • Default: Use the /image/generate or /structured_prompt/generate endpoints for state-of-the-art input understanding and detailed prompt creation.
    • Open: (Coming Soon) Use the /image/generate/open or /structured_prompt/generate/open endpoints, powered by our open-source VLM, enabling flexible on-prem deployment.
  • Capabilities:
    • Generate new images from text or reference images.
    • Refine or recreate previous generations using their structured_prompt.
    • Optionally generate only the structured_prompt for inspection, editing, automation, or hybrid workflows.

Legacy V1 Image Generation Pipelines:

These endpoints offer specific pipelines but lack the structured prompt control of the latest generation.

  • Use the /text-to-image/base, /fast, or /hd endpoints for various image generation needs (quality, speed, resolution). Supports Guidance Methods.
  • Use the /text-to-vector/base & /fast endpoints (Beta) to generate illustrative vector graphics. Supports Guidance Methods.
  • Utilize the /reimagine endpoint to generate new images preserving structure and depth from a reference image.
  • Enhance your prompts with the /prompt_enhancer route to create more detailed descriptions for V1 generation.

Dive into the Docs

Tailored Generation:

The Tailored Generation APIs allows you to manage and train Tailored Models that preserve and faithfully reproduce your visual IP (photos, illustrations, vectors). You can train Tailored Models through our Console or implement training on your platform via API. Check out our Console here.

Training Process Overview

To train a tailored generation model:

  1. Create a Project: Use the /projects endpoint to create a project and define your IP type and medium.

  2. Create a Dataset: Use the /datasets endpoint to create a dataset within your project.

  3. Generate an Advanced Caption Prefix

  4. Upload Images: UUpload images using the /datasets/{dataset_id}/images or /datasets/{dataset_id}/images/bulk endpoints (minimum resolution: 1024x1024px).

  5. Prepare Dataset: Review auto-generated captions and use the /datasets/{id} endpoint to set status to 'completed'.

  6. Create Model: Use the /models endpoint to create a model, selecting either the "light" or "max" training version (Fully Automated mode) or the training parameters (Expert mode)

  7. Start Training: Initiate training using the /models/{id}/start_training endpoint. Training takes 1-3 hours.

  8. Monitor Progress: Check training status using the /models/{id} endpoint until status shows 'Completed'.

  9. Generate Images: Once trained, your model can be used in multiple ways:

    • Use the /text-to-image/tailored/{model_id} endpoint for text-to-image generation
    • Use the /text-to-vector/tailored/{model_id} endpoint for generating illustrative vector graphics
    • Use the /reimagine/tailored/{model_id} endpoint for generating using a structure reference image
    • Access through the Bria platform interface

Dive into the Docs

Ads Generation

Bria's Ads Generation set of APIs streamlines the process of creating, managing, and customizing advertising campaigns, providing powerful tools for scalable and efficient ads and campaign generation.

These tools include:

  • /create: Generate ads using Template IDs and content inputs, enabling rapid creation of professional-grade ad variations.

  • /brands: Retrieve a list of brands along with their associated logos, colors, and fonts, providing the building blocks for brand-aligned campaigns.

  • /templates: Access a list of templates, including their IDs, names, number of scenes, and a preview of the first scene.

  • /templates/{template_id}: Fetch detailed JSON data for a specific template, including all its elements, enabling precise customization and adaptation.

Dive into the Docs

Product Shot Editing

Bria offers a variety of solutions for products and apps in the eCommerce industry, which allow to enhance product imagery for online stores and product catalogs.

These solutions include:

  • /cutout: Create clean and professional product cutouts.

  • /packshot: Generate professional-standard packshots with customizable background colors and optimal product placement.

  • /shadow: Add consistent and customizable shadow effects, supporting regular and floating shadows for product images with transparent backgrounds.

  • /lifestyle_shot_by_text: Enrich product images by placing them in various environments using textual descriptions, enhancing appeal and providing context.

  • /lifestyle_shot_by_image: Generate product images in enriched environments using a reference image for inspiration, allowing for creative and unique presentations.

  • Automotive Product Shot Editing: Dedicated support for automotive workflows, including vehicle-specific features such as realistic reflection generation, tire and wheel refinement, environmental effects (e.g., fog, dust, snow), and harmonization presets.

Dive into the Docs

Image Editing

Bria supports both isolated and registration-required routes for editing images:

  • /enhance - Generate richer details, sharper textures, and enhanced clarity while upscaling the resolution of an image.

  • /increase_resolution - Upscale any image.

  • /expand - Expand the image beyond its original borders, adjusting it to different aspect ratios.

  • /remove_background - Remove the background of an image.

  • /blur_background - Create a blur effect on the background of an image.

  • /replace_background - Generate new backgrounds for an image.

  • /erase_foreground - Erase the foreground from a provided image.

  • /erase - Erase elements from an image using manually created or auto-generated masks.

  • /gen_fill - GenFill generates elements in a specific region of an image, using a manually created or auto-generated mask and a textual prompt.

  • /objects/mask_generator - Generate masks for objects in an image.

Registering Images before editing:

  • Use the /register endpoint to send a new image to Bria without being uploaded to the database. This is the recommended method to onboard images to Bria when needed.

With the response from the Upload/Register Routes (visual_id), the following routes can be used without any particular order:

  • /person/info (soon to be deprecated) - By using the /person/info route, you can understand the possible changes for every person in the image.
  • /person/modify - Modify person-related elements in the image. In order to use this route you should first use the route /person/info.

Dive into the Docs

Previous page
Next page