Selecting the right model
Seelab's ambition is to make AI-powered image generation easy and accessible to everyone. For instance, Seelab offers a library of public models (or styles), so you can easily create stylized images.
To select a model, simply click on it in the left side of the interface.
The selected model will appear above the prompt window (see the example above with the "Modern Pop Art" model).
Seelab regularly adds new public models. They are grouped into 5 different categories.
*Photography: a selection of photo-realistic models inspired by famous film stocks or photography styles.
*Illustration: from black & white sketches to comics, over 15 different illustration styles.
*Digital: from Pixel art to vectorized style, including Cel Shading and Low Poly.
*Artistic: from Painting effects to Claymation, an astonishing selection of artistic styles.
*Generalist: if you don't have a specific style in mind, you can choose a generalist model and let yourself be surprised.
If you train your own model from your images, it will appear under the "Project Models" tab. You can also share it with other users in your company. For more information, refer to the "Create Your Custom Model" category.
Each model offered on Seelab.ai is unique. Beyond style, each model will respond differently to user prompts and will therefore be more or less suitable for certain types of images.
For example, some illustration models may struggle to create character images without generating artifacts and will focus more on style. Similarly, the rendering of objects may also be impacted by the model used. Example: the "Kodachrome Vintage" model tends to generate cars from the 70s/80s unless specified otherwise in the prompt.
In some cases, it may be easier to generate an image using the "Photo Realistic" models, and then stylize it by using the "Reference Image" feature and another model.
Above, an example of the same prompt rendered with different Seelab models.
Prompt: "A robot character, pink flowers, dark background, complex cybernetics"
How does it work?
To select a model, simply click on it in the left side of the interface.
The selected model will appear above the prompt window (see the example above with the "Modern Pop Art" model).
Public Model Categories
Seelab regularly adds new public models. They are grouped into 5 different categories.
*Photography: a selection of photo-realistic models inspired by famous film stocks or photography styles.
*Illustration: from black & white sketches to comics, over 15 different illustration styles.
*Digital: from Pixel art to vectorized style, including Cel Shading and Low Poly.
*Artistic: from Painting effects to Claymation, an astonishing selection of artistic styles.
*Generalist: if you don't have a specific style in mind, you can choose a generalist model and let yourself be surprised.
Custom Models
If you train your own model from your images, it will appear under the "Project Models" tab. You can also share it with other users in your company. For more information, refer to the "Create Your Custom Model" category.
Each Model is Unique
Each model offered on Seelab.ai is unique. Beyond style, each model will respond differently to user prompts and will therefore be more or less suitable for certain types of images.
For example, some illustration models may struggle to create character images without generating artifacts and will focus more on style. Similarly, the rendering of objects may also be impacted by the model used. Example: the "Kodachrome Vintage" model tends to generate cars from the 70s/80s unless specified otherwise in the prompt.
In some cases, it may be easier to generate an image using the "Photo Realistic" models, and then stylize it by using the "Reference Image" feature and another model.
Above, an example of the same prompt rendered with different Seelab models.
Prompt: "A robot character, pink flowers, dark background, complex cybernetics"
Updated on: 30/10/2024
Thank you!