Create images from text prompts

Go to the Text to Image feature in the Firefly web app, hover over any image in the gallery, and select View.

Select Firefly Image 2 or Firefly Image 3 from the Model dropdown menu in the General settings section.

If you’re using Firefly through an organization-owned account with Custom Models and have the entitlement to generate them, you’ll see Custom Models in the dropdown list. Select one to create on-brand assets for your organization. Contact your Adobe administrator to request access to generate Firefly Custom Models.

Select the aspect ratio from the following options:

• Landscape (4:3)
• Portrait (3:4)
• Square (1:1)
• Widescreen (16:9)

In the Content Type section, select Photo or Art. Turn on the  Auto mode to have Firefly automatically select a content type for you and set the photo settings.

When you select Photo as the  Content Type, manual camera controls are enabled under Photo settings. 

• Aperture allows you to control the background blur in your image by setting a specific f-stop.
• Shutter speed allows you to control the sharpness of your image.
• Field of view allows you to use a specific camera lens in your generation.

In the Composition section, upload a reference image or select one from the gallery. This reference image will be used to influence and match the composition of the generated image variations. 

If you’re using Firefly with an organization-owned account that includes the Object Composites entitlement, you’ll have the option to upload an image of an object and create a composite scene with it.

Use the Strength slider to adjust the strength of the outline and depth adherence when generating the image variations. 

In the Styles section, use the Visual intensity slider to modify the intensity of the photo's visual attributes.

• For Content Type set as Photo, you can transform them to be more realistic (Left) to surrealistic (Right).
• For Content Type set as Art, you can transform them from digital art (High) to something more illustrative (Right).

Use the Strength slider to adjust how closely the generated image variations should align with the style reference image.

Use the Reference option to create an image based on the look and feel of a reference image. Either upload an image under Style > Reference or use Browse gallery to select one from the image gallery.

You can even combine reference images with Effects to create fresh and unique aesthetics.

Apply Effects and specify Color and tone, Lighting, and Camera shot type. The generated image will combine the prompt you enter and these specified controls.

Enter a detailed prompt.

For suggestions to complete your prompt, turn on the Prompt suggestions toggle. Then, write at least two words (English only) and pause for the suggestions to appear. Select a prompt to use it, select the Refresh   icon for more suggestions, or hover over a prompt to select the Report   icon and share feedback. (Desktop only)

Once the prompt is ready, select Generate.

Select the image you like from the generated results. Alternatively, try rewording your prompt to get something closer to what you want, or select Generate to generate more image variations.

Hover over an image for more options.