Layla v5 supports generating images with Stable Diffusion models.
In Layla, you can generate images with several methods:
Using your device itself (without connecting to any external providers)
Connecting your phone to your PC
Using Layla Cloud
Whichever method you choose, you must enable the Stable Diffusion mini-app in Layla:
Using your device itself
The image generation will be done by your device's (phone or tablet) CPU. Layla provides a variety of Stable Diffusion models pre-built into the app. They can be downloaded by going into the Stable Diffusion mini-app:
Tap the blue "cloud download" button to download the model (this may take some time as the models are pretty large). Tap the model tile at the top to choose from different models.
Selecting a model and downloading the files, you will be allowed to input your prompt and other settings to generate an image.
Connecting to your PC
If you have a PC, you can install a popular Stable Diffusion WebUI built by Automatic1111: https://github.com/AUTOMATIC1111/stable-diffusion-webui
Setting up Automatic1111-sd-webui is beyond the scope of this tutorial. You can follow the ReadMe file in their GitHub link or follow various other tutorials on YouTube.
After you set it up and configure the API, you can connect to it in Layla by going to the Inference Settings app. Scroll down to the Image Generation settings:
Tap "Add Custom Model". You will be able to configure your API settings:
You can find your PC's IP address through various means such as your router.
After you configure your PC, it will be available as a Stable Diffusion Model to choose from when you generate images:
Using Layla Cloud
You will notice some models have a butterfly symbol at the top right. This means these models are provided by Layla Cloud, and requires you to purchase a subscription in Layla Cloud app. All other models are locally generated on your phone.
These models are provided by Layla Cloud and requires a corresponding subscription to use. They provide fast, seamless image generation.
Allowing your characters to send you images during chatting
Lastly, you can allow your characters to generate images while chatting. This feature is only available for custom characters.
In the character creation screen, you can configure the image generation settings:
You can choose your stable diffusion model for this specific character.