The Create/Edit Caption dialog supports an AI Source. When Source is changed from Manual to AI, the dialog exposes AI generation controls that can create a caption or improve existing caption text.
If the Caption Body is blank, the action appears as Generate Caption. If caption content already exists, the same action can appear as Improve with AI.
Select Show Settings to display the user-level AI generation options. These are content-generation preferences, not provider API configuration.
You can upload an optional image reference. The browser creates a preview of the selected image and includes the image data in the caption-generation request.
Use Custom Instructions to tell the caption generator about required messaging, products, offers, campaign details, or other guidance.
The AI caption dialog provides:
Select the format that best matches the content for which the caption will be used.
Source-defined Tone choices include:
The current dialog provides language choices including:
Use Char Limit to guide the maximum requested caption length. The default AI setting is 2,200 characters.
Enter relevant Keywords such as products, topics, campaign themes, or audience terms. The interface provides an example similar to fitness, wellness, summer...
The AI request asks for three caption options. When generation succeeds, the current Captions Workspace dialog places the first returned caption into the Caption Body and can report that multiple captions were generated.
You can edit the resulting text manually before saving it.
Always review generated content for factual accuracy, brand fit, offer details, spelling, tone, platform suitability, and any required disclosures before using it.
AI caption generation can depend on the caption credits and AI-generation access available to your account. The normal user does not configure AI provider keys, global AI models, or service credentials as part of this workflow.