Skip to content
Candy AI
Free accessFree uses

Experiences

Create a new picture with the Candy AI image generator

Choose the intended companion, decide what should be visible and use the supported picture tool. Outfit, pose, surroundings, lighting and viewpoint describe the composition. A conversation can explain why the characters are there, but the picture still needs its own visual instructions and a separate cost check.

On this page

Review the request before submitting it

Confirm the character when similar names appear, then choose the image action available to your account. Check the requested quantity and displayed token cost rather than guessing a universal price from your balance. An image-generation entitlement and payment for the individual request can be different conditions.

  1. Identify the companion in the supported image tool.
  2. Specify the clothing, pose, place, light and framing.
  3. Check the requested quantity and charge in the confirmation.
  4. Review the completed picture before making a revision.

Keep all visible details in the same scene

A waist-up portrait beside a greenhouse window can include one pose, outfit and clear light source. Asking for a distant landscape and an extreme facial close-up at once makes the intended composition ambiguous. Start with the frame that matters and add details that can coexist within it.

Compare the result with the original intention. If the pose is the biggest mismatch, clarify it instead of simultaneously changing every other part. Reusing identical wording can still produce variation; it does not guarantee the same image, flawless anatomy or perfect visual continuity.

Enable supported image mode through Ask

In an eligible conversation, open Ask and inspect Image Generation Toggle. Turning it on enables the offered image mode; read the charge attached to any actual request. It does not convert generation into ordinary written messaging or remove the need for sufficient media tokens.

Turning the mode off affects that picture workflow. Calls, clips, packs and other selected operations retain their own conditions. If the toggle is absent, check the character and chat mode rather than assuming all profiles should show an identical image menu.

Distinguish fresh generation from editing a photograph

This workflow requests a new companion image from the selected character and description. It is not an uploaded-photo editor or a promise of identity transfer. An input accepted by some other tool does not establish that arbitrary photographs can be supplied here.

Review completed material using its generated-media controls. Any tool that explicitly accepts a picture has its own requirements, and you must hold the necessary rights and consent. Saving the output or a profile reference does not remove content restrictions or grant permission to reuse another person's likeness.

Your questions

How do I estimate the image request cost?

Use the price attached to the chosen tool and quantity, then inspect the final confirmation. A monthly balance or token-pack size is not a fixed count of pictures. Check again before another attempt because access to the generator does not by itself pay every request.

Which control turns on pictures in a chat?

Look under Ask for Image Generation Toggle in a supported conversation. Review both that mode setting and the cost of the selected request. A different character or mode may not expose the same control, so do not assume an absent switch is a payment failure.

Can I supply a photograph just because I have image access?

No. The selected tool must explicitly accept the input, and access does not supply another person's consent or likeness rights. A new-character image generator should not be treated as an undocumented photograph editor or identity-transfer service.

Is this an editor for my existing photo?

The request described here creates a new companion picture. Editing an uploaded photograph is a different capability and is not established by this access. Follow the accepted inputs of the actual tool and use only material permitted by the content and consent rules.