OnlyFans Promo Photos: Prepare Your Next Post Without Another Shoot
Use an existing photo or make a new promotional look in StarSmith. Follow the selfie-to-download steps and fix common image problems before posting.
You've finished making content for OnlyFans, but you still need a public photo. Before getting ready for another shoot, look for a usable picture from the work you've already done. If you need a different look, StarSmith can generate a promotional image from your selfie and a visual reference or prompt.
This walkthrough takes you from that choice to a downloaded image and caption. It is for adult creators preparing off-platform promotion; a generated picture isn't a substitute for the real content you sell on OnlyFans.
Choose an existing photo or a new generated look
Start with the post you're trying to make.
| You need… | Use… |
|---|---|
| A preview of a particular paid set | A suitable image from that actual set |
| A fresh introduction to your public creator account | An unused real photo, or a generated look where AI images are accepted |
| A submission for a community that requires original photography | A real photo meeting that community's rules |
| A different crop of an image you already like | A separate export of the existing photo |
The distinction matters to creators choosing what to give away publicly. In this Reddit discussion about selecting promo pictures, the poster was weighing photos for promotion against photos to keep on the paid page. The comments offered different approaches, not a universal amount to reveal.
For an existing photo, check the expression, framing, and background at the size people will see it. A good crop may be all you need. Keep the original and save the public version separately.
For a generated introduction, write a brief such as: “A relaxed, fully clothed portrait for my public account, with my face clear and a playful expression.” That's our example brief, not a tested customer campaign. If you're advertising a real release, return to its actual photos instead.
Check whether that placement accepts AI images
Choose the platform or subreddit before spending credits. A community may accept adult creator photos while excluding AI-generated images, edits, or seller accounts. The subreddit guide covers how to establish those conditions; the Instagram guide covers that platform's content and recommendation restrictions.
Fansly prohibits photorealistic AI-generated content. Don't use the generated looks in this tutorial as Fansly posts or FYP previews. Adding an AI label doesn't create an exception to that prohibition.
If the chosen placement doesn't accept the image type, use a real photo or find an eligible placement. Regenerating or changing the caption won't fix that mismatch.
Make a promotional image in StarSmith
1. Open Create and select your selfie
In Create, upload or select a photo under Selfie. Use a clear image of yourself with your face visible. A blurry face, strong filter, or obstruction makes it a poor starting point when likeness matters.
Your selfie supplies identity and face. The reference you select next supplies the look. These are different inputs: putting your selfie into the reference slot won't serve the same purpose.
2. Choose a reference, write a prompt, or use both
If you don't know what to make, open Inspiration and select Use it on a look you like. It opens Create with that Inspiration as the reference. You can also upload a reference you have the right to use.
Pick the visual details you actually want: composition, pose, clothing, mood. For a creator whose public photos are relaxed and conversational, a dramatic costume reference may give you more work to correct than a simple at-home look.
You can use one reference, a written prompt, or both. With a reference selected, the prompt can add a specific adjustment. Without a reference, it needs to describe the direction.
For example:
A fully clothed at-home portrait in a black top, with soft window light, a playful expression, and an uncluttered background.
That's an original prompt example to adapt. Start with the details that matter for the intended post; adding several incompatible poses or outfits makes the result harder to judge.
3. Pick the image size and select Generate
Choose the available size that suits the placement. Check the framing you'll need—portrait, square, or taller—before generating. Then select Generate.
Each successful image uses 4 Credits. A successful replacement or another version uses credits too. Generation isn't free with registration; check the current plans before starting.
If Generate is unavailable, check that your selfie has finished uploading and that you've provided either a Reference or Prompt. Read any account or credit message shown alongside the controls.
4. Review the result before downloading
Inspect the result at full size and at the smaller size of a social feed:
- Does the face look like you? Check body proportions and visible features as well; don't assume every tattoo, skin detail, or accessory is preserved.
- Are the hands, clothing edges, jewelry, and reflections coherent?
- Does the frame show only what you intended to make public?
- Would your planned caption imply a real location, shoot, or matching paid collection that doesn't exist?
Use the result's Download control when you have an image you want to use. Completed images also stay in Gallery → Creations, so you can return to the result later. The image-creation docs explain the inputs, storage, and credit rules.
For a visual reference, the Instagram article includes eight actual StarSmith outputs from one source image. The demonstration source is itself AI-generated. Those examples show output variation, not a real creator's revenue or measured time saving.
What to change if a generated photo looks wrong
| Problem you can see | What to change first |
|---|---|
| The face doesn't represent you well | Review the selected selfie for clarity and obstructions. The More Like Me control provides an identity-refinement path; inspect the new result too. |
| The face is recognizable, but the clothes or setting are wrong | Change the reference or the specific prompt detail responsible for that look. |
| Hands, accessories, or reflections look wrong | Simplify the problematic pose or scene before making another version. Don't publish a defect just because the rest looks good. |
| The crop cuts off the part you wanted to show | Reconsider the chosen size and composition. If a crop of the downloaded image solves it, another generation may be unnecessary. |
| The picture is attractive but doesn't fit your account | Choose a reference closer to the style you want people to associate with you. |
| The subreddit rejects AI images | Use an eligible real photo. This is a placement problem, not a generation problem. |
These are practical troubleshooting choices, not promises that a retry will fix everything. If the changes keep taking longer than a quick real photo, use the camera for that post.
Add the caption and publish it yourself
On the completed result, choose Generate caption. Use New caption if you want another suggestion. StarSmith selects from a human-reviewed general collection; it doesn't analyze the picture or tailor the sentence to a specific platform.
Copy or edit the line to fit what the photo actually shows. The promo-caption guide gives examples of that last edit, including a generated image with no matching paid set.
Save the picture and caption together, follow the placement's AI disclosure requirements, and publish in the destination app. StarSmith doesn't log into your social account or post for you.
Use Smith when getting started keeps slipping
Smith can bring an Inspiration to Telegram at your saved time. You can use it to start a generation, review the result when it arrives, and request a caption. That gives you a concrete starting point when “make a new promo photo” keeps getting postponed.
After posting, check in with Smith through the web or Telegram; they share your daily status. This records your confirmation, not clicks or subscriptions. The Telegram guide explains connection and reminders.
Max users can separately opt into Autopilot for daily image-and-caption preparation. It uses credits, and you still review and publish the result yourself.
For your next few posts, compare the whole preparation time with your usual approach: choosing the look, generation, review, retries, caption, and export. That tells you whether the workflow saves you time. Timing only the generation would leave out the work you still have to do.
Your next public promo image
See what you can make from your selfie
Choose a reference, generate a look, and review the result before posting.