Set up the image request
Open the generator's Image mode and choose Existing Character or New Image. Work out the subject first, then choose the method and framing that fit it.
- For Existing Character, select the companion you want to depict; use New Image for a separate subject.
- Choose parameter-based generation or write a text prompt for the scene.
- Select Portrait, Landscape or Square and decide what must fit inside the frame.
- Check the image count and displayed token total before sending the request.
- Select Generate Image only when the request is ready. Inspect the result before spending on another version.
Describe what belongs in the frame
Describe the visible subject, place and composition rather than a long backstory. A brief might read: a fictional adult marine biologist in a green rain jacket, holding a field notebook at a quiet harbor, three-quarter portrait, soft overcast light.
Choose one visual treatment. If you want anime or cartoon styling, say so without combining conflicting looks. When using a selected character, let the new prompt establish the setting instead of trying to replace every defining feature.
- In a portrait, decide how much of the subject should appear and which expression matters.
- In a wider scene, give the surroundings a clear role, such as a harbor railing beside the character.
- If you may animate the image later, leave room for the action instead of cropping tightly around the face.
Inspect the picture and find it again
Check the face, hands, clothing, objects and framing. When a detail misses your brief, name the change you want for the next request. Keep the original wording so you can compare revisions instead of starting over from memory.
Open Recent Generations and use the image filter to look for a result. Check whether My or Everyone is selected. Those view names help you navigate; they do not establish a particular output's visibility, so check that separately before sharing.
Image generation is listed at 2 tokens. Review the total for your selected image count rather than assuming every batch has the same charge. Unlimited messages on a subscription do not make image generation unlimited.
Common questions
How does DarLink's AI image generator work?
Choose an existing character or a new image, set the scene with parameters or a text prompt, and select the framing. Check the image count, access notice and charge before choosing Generate Image.
Can I put a character I already made into a new scene?
Yes. Select the existing character and describe the new surroundings or composition. Keep the instructions about the picture focused so they do not contradict the character's defining features.
Can I generate without writing a prompt?
Parameter-based generation is also offered. Use those controls when they describe what you need, or a text prompt when the scene calls for your own wording.
What framing options are available?
Portrait, Landscape and Square describe the image's shape. They are not pixel specifications; check separately when you need a particular resolution or downloaded file format.
Does Ultimate remove the image token charge?
No. Ultimate's unlimited allowance applies to messages. Image requests still carry a listed token cost, so check both the total and your available balance.
Can I use the generator to edit an uploaded photograph?
Creating a new character image is not the same as retouching an outside photo. Check the available upload and editing controls for the task you need before importing a file. Do not assume face swapping or other transformations are supported.
Will an image request produce an animated clip too?
No. Image and video requests are separate. To add movement, use the video workflow and review its own access requirements, settings and token charge.
For an account question, email the DarLink support team.