Add Editable Text to an AI Image

Reserve space before you add words

Choose an image with a calm area large enough for the intended message. Our mug occupies the lower portion of the square, leaving a relatively even background above it. That space gives the headline a place to sit without covering the handle or rim.

Write the real headline first. A composition that fits four words may not fit a long date, venue and offer. Put supporting detail in the accompanying caption or another deliberate text area; shrinking everything until it fits usually weakens the message.

Use Text in the canvas toolbar

Choose Text in the bottom toolbar, then click the intended position on the canvas and enter the headline. Finish editing to select the new object. It should expose text controls rather than image-editing controls.

In this session, asking the agent to create editable text returned an explanation that the action must be performed with the Text tool. We therefore added the headline manually. Do not assume that a chat request executes every action available in the canvas.

Make your first mug

Adjust the text object, not the picture

Select the headline and use its Fill, font and size controls. We used a white fill over the slate background and reduced the size until the full line fit inside the image. The correct size depends on your composition; a number copied from another board is only a starting point.

Keep some space between the words and the image edge. Check the longest word and any descenders, accents or punctuation. A bright highlight behind one part of a white headline can make that part disappear even when the rest reads well.

Prove the words are still editable

Double-click the text object and correct a word or replace the line. Finish editing, then read the entire headline again. This simple check catches accidental misspellings and confirms that you are editing text rather than asking an image model to redraw letters.

If the image itself already contains words, adding a text object does not remove them. Start with a text-free image or deliberately repair the original lettering first. Otherwise two versions of the message may overlap.

Check at the size your audience sees

Zoom out until the square is close to its expected viewing size. Read the headline without leaning in, then inspect whether the mug still has enough visual weight. A large editing view can hide weak contrast and crowded margins.

Keep the accepted image and editable text in the project. For another language, replace the text and review its line breaks rather than assuming the same width will work. The original artwork can stay unchanged if the composition still supports the translated copy.

Export the intended deliverable

An image source file and a canvas composition are different deliverables. Ciyo's active-asset Source file option downloads the generated image; it does not promise to include a separate headline placed on the canvas. In our example, the downloaded source remained the text-free mug image.

If a teammate needs to continue design work, hand over the source image and exact headline separately with placement notes. If you need one flattened campaign graphic, use the appropriate composition export and open the resulting file to verify that all intended objects are present. Do not approve an export solely because the canvas looked correct.

• Export an image with its prompt and model record