Image Generation in Tess AI
Two main paths: templates (pre-configured and user-created) and Image CoPilot (total customization).
Templates
Available in the community Marketplace. They allow you to generate images following a predefined training line. Examples: logos, thematic illustrations, upscale, etc. Users can create their own templates through the individual plan.
Image Prompts
Prompts are commands that guide the AI in creating the image. The more details provided, the better the result. It is recommended to use prompts in English. Examples and tips are available on the Tess AI blog and in the Help Center.
Showcase
A public feed with images generated by users. It serves as inspiration and an image repository. When you open an image, you can view the prompt, the AI model, the size, and other information.
Image CoPilot
Offers a high degree of personalization. It provides a field for prompts, selection of AI models (Google, Stable Diffusion, Degr, Tess, etc.), and advanced settings (style, lighting, mood, size, etc.).
Prompt Tips
Use the "image prompt" template or the Chat CoPilot to generate prompts. The recommended structure is: main object + attributes + details + style + lighting + angle + engine + artist, etc.
Advanced Settings
They vary depending on the AI model. Options include art style, lighting, mood, size, inference (number of generation steps), AI scale (creativity level), similarity strength, Tess Alquimia (image refinement), negative prompt (elements to avoid), and Seed (a parameter to control randomness).
Flux Schnell and Tess AI V4 allow unlimited image generation at no extra cost!
Image Editor (Canva): Allows you to edit images generated by the AI (add logos, text, shapes, etc.). It creates a new version of the image without altering the original. You can export the design file for external use.InPaint and OutPaint
InPaint allows you to modify specific elements of the image using masks. OutPaint expands the image by adding content around it.
Discover all this on Tess AI!