BAGEL is a high-quality text-to-image model with excellent prompt adherence and creative capabilities. Supports both text-to-image and image-to-image generation. Supports thought tokens for enhanced generation quality.
Added May 26, 2025
Approx. Price
$0.170 per image
Model Type
both
Settings
Generation controls available for this model.
Images Per Run
Up to 1
Output images
Input Images
Supported
Reference/edit images accepted • Route max 30 MB
Output Sizes
Enable Safety Checker
Default
Yes
Number of Images
Default
1
Resolution
Default
1024x1024
Options (1)
1024x1024 (Square (1:1))
Seed
Default
N/A
Seed for reproducibility. Leave empty for random.
Use Thought Tokens
Default
No
Options (2)
On, Off
Enable thought tokens for potentially improved generation quality. Increases generation time and cost by 20%.
Benchmarks
Benchmarks
Human preference benchmarks sourced from Artificial Analysis.
Text to Image
#145 / 151
ELO
909.0
Appearances
3,598
95% CI
-9/9
Image Editing
#67 / 70
ELO
916.0
Appearances
6,948
95% CI
-9/9
Release Date 2025-05 · Matched as Bagel
Artificial Analysis APIExamples
Loading examples…
Related image models
Compare BAGEL with similar models from the same provider or model family.
Seedream 5.0 Pro
bytedance/seedream-v5.0-proByteDance Seedream 5.0 Pro for high-quality text-to-image generation and image editing. Upload images to switch to edit mode.
Seedream 5.0 Lite
seedream-v5.0-liteByteDance Seedream 5.0 Lite for text-to-image and image editing with strong prompt adherence, improved typography, and up to 4K output. Upload an image to switch to edit mode. Minimum total pixels: 3,686,400.
Seedream 5.0 Lite Sequential
seedream-v5.0-lite-sequentialSeedream 5.0 Lite Sequential generates consistent multi-image sets in one run, with unified style, palette, and character continuity. Upload images to switch to edit-sequential mode. Supports up to 4K output.
Seedream 4.5
seedream-v4.5Seedream 4.5 is the latest image model with improved quality. High-quality results with sizes up to 4096x4096. Minimum total pixels: 3,686,400 (e.g., 1920x1920 or 1664x2496). Max input image size: 10MB.
Seedream 4.5 Sequential
seedream-v4.5-sequentialSeedream 4.5 Sequential generates multiple consistent images with character and object consistency. Maintains unified palette, lighting, and style across outputs. Supports up to 4K resolution. Max input image size: 10MB.
Seedream 4.0
seedream-v4Seedream 4.0 is a state-of-art image model. High-quality results with sizes up to 4096x4096. Max input image size: 10MB.