Settings

Configure per-project options that control how your images are generated.

Generation Model

Choose which image generation model to use for your project. Different models offer different trade-offs between cost, speed, and quality.

ModelCreditsSpeedBest for
FLUX 1.1 Pro4010-15sBest quality, photorealism
FLUX.1 Dev308-12sGreat quality, slightly faster
FLUX.1 Schnell101-2sQuick previews, iteration
Recraft V34010-15sVector art, icons, brand design
Seedream 4.53010-15sPhotorealism, text rendering
Ideogram V34010-15sTypography, logos, labels
Imagen 44010-15sState-of-the-art realism
WAN 2.73010-15sArtistic, detailed output
Qwen Image2015-20sComplex scenes, text rendering

Auto Mode

Auto is the recommended default. When enabled, the AI analyzes each item's complexity during prompt generation and selects the most appropriate model:

  • Simple items (icons, badges, flat illustrations) → FLUX Schnell (10 credits)
  • Standard items (food, products, everyday objects) → FLUX Dev (30 credits)
  • Complex items (detailed scenes, photorealistic subjects) → FLUX Pro (40 credits)

This can significantly reduce costs for projects with a mix of simple and complex items, without sacrificing quality where it matters.

Remove Background

Toggle automatic background removal for generated images. When enabled, BiRefNet v2 removes the background from every image to create a transparent PNG. The original is also saved.

  • Default: Enabled
  • Cost impact: +40 credits per image when enabled
Settings can also be changed through the Brainstorm chat. Ask the AI to "switch to FLUX Schnell" or "turn off background removal" and it will update the setting for you.

Coming Soon

Additional settings are planned for future releases:

  • Output resolution — Choose from standard sizes (currently fixed at 1024×1024)
  • Auto-retry — Automatically retry failed image generations