AI Image Models
ox0.in runs open-source diffusion models developed by Black Forest Labs. These models generate high-quality images from text prompts and are known for strong text rendering, anatomical accuracy, and artistic versatility.
FLUX.1 [dev]
Best quality Balanced speed Open weights
FLUX.1 [dev] is an open-weight guidance-distilled model designed for high-quality image generation. It offers an excellent balance between output quality and inference speed, making it the default choice for most users on ox0.in.
| Attribute | Value |
|---|---|
| Developer | Black Forest Labs |
| Architecture | Flow-based rectified transformer (12 billion parameters) |
| License type | Open weights for non-commercial and commercial use under the FLUX.1 [dev] Non-Commercial license; commercial licensing available from Black Forest Labs |
| Best for | Photorealistic portraits, detailed illustrations, complex scenes with multiple subjects, readable text, cinematic images |
| Speed | Moderate — typically 10–40 seconds depending on queue length and prompt complexity |
| Output resolution | Up to 2 megapixels (1024×1024, 1024×1536, 1536×1024) |
FLUX.1 [schnell]
Fastest Apache 2.0 Local-friendly
FLUX.1 [schnell] is the fastest openly available FLUX.1 model. It is released under an Apache 2.0 license, which makes it suitable for personal projects and commercial use without additional licensing. The trade-off is slightly less refinement compared to [dev], especially for complex scenes.
| Attribute | Value |
|---|---|
| Developer | Black Forest Labs |
| Architecture | Flow-based rectified transformer, optimized for 1–4 inference steps |
| License type | Apache 2.0 (fully open) |
| Best for | Quick drafts, rapid iteration, simple scenes, icons, concept thumbnails, social media images |
| Speed | Fast — typically 5–20 seconds depending on queue length |
| Output resolution | Up to 2 megapixels (1024×1024, 1024×1536, 1536×1024) |
Quick comparison
| FLUX.1 [dev] | FLUX.1 [schnell] | |
|---|---|---|
| Image quality | Excellent | Very good |
| Speed | Moderate | Fast |
| Text rendering | Strong | Good |
| Complex prompts | Better | Good |
| License | Open with usage terms | Apache 2.0 |
| Best use | Final production images | Rapid prototyping |
Aspect ratios and resolutions
- 1024 × 1024: Square format ideal for social posts, avatars, icons, and product shots.
- 1024 × 1536: Portrait format ideal for mobile wallpapers, character portraits, and vertical social content.
- 1536 × 1024: Landscape format ideal for desktop wallpapers, banners, and wide scenes.
Influence parameter
The Influence slider controls how closely the output follows your prompt. Higher values make the model adhere more strictly to every word in your description. Lower values give the model more creative freedom.
Responsible use
These models are powerful tools. Do not use them to generate misleading, harmful, hateful, non-consensual, sexually explicit, or illegal content. Always follow our Terms of Service and respect the intellectual property rights of others when using generated images commercially.
Generate with these models →