Create images and write post text
Make a visual from an idea, prepare captions and keep the difference between a prompt and a public post clear.
Browse guides
On this page
Not every idea needs a video. Use an image for a product visual, illustration or poster, and use text when the words are the content.
Generate an image
- Open Create a post and select Image.
- Describe the subject and composition. Add the setting, lighting, style and any words you want inside the image.
- Add reference images if the selected model offers them. Use references you have permission to use.
- Review the model, settings and cost or access message.
- Choose Generate content, then inspect the result before sharing it.
The image catalog currently includes GPT Image 2.5 Flare. Availability can change, so the selection shown in your studio takes precedence.
Use a starting format or style
Starting formats such as talking objects, cute animals, absurd memes, quote posters and flat lays help fill in an initial idea. They are editable starting points. Choosing one does not generate an image by itself.
Check the full prompt after choosing a format. Change the details to fit your own idea. A visual style helps communicate the look you want, but it cannot guarantee an exact output.
Prepare text
Select Text to generate a written post from a topic. Tell the model who it is for, what you want to say, the language and the tone. The text catalog currently includes Gemini 3.5 Flash.
For example:
Write a short, friendly Instagram caption about a quiet morning coffee. One sentence, no hashtags, no claims about health or productivity.
Read the result and make it your own. Check facts, names and any claims before you publish.
Keep prompts and captions separate
A prompt tells AI what to make. A caption is the text someone reads beside your post. Changing a caption does not change the picture or video, including text already visible inside it.
The video creation form has Write with AI and optional instructions for post text. The publishing review also lets you review captions and prepare different wording for individual accounts. Requesting a new AI description is a paid text action when the app says so.
Work with the finished image
Open the result or find it in Library. You can download it, use it as a source for a video or prepare a publication. Where Remove background is offered, you can make a cutout and choose a replacement background. Review the saved result before using it.
Image and text posts cannot go to every destination. X accepts images and text, while YouTube and TikTok require video. Connecting social accounts lists the supported formats.
Still stuck? Contact support or check access and troubleshooting.
Your next idea starts here.
Make a video, bring an image to life or try something unexpected.