Generation and Media Problems
Recover from image, selfie, Try-On, video, avatar, world, and voice failures without repeating paid actions unnecessarily.
Updated July 25, 2026
Generation is not always immediate. An image, video, avatar, world, or voice can move through queued, processing, completed, or failed states.
Before you retry
- Read the exact message shown in the app.
- Wait if the item is still queued or processing.
- Check the relevant character, chat, image, or video gallery.
- Confirm that you selected the correct character and workflow.
- Check your subscription status, PORTI balance, and current access.
- Retry once with one small change.
A timeout does not always mean the generation stopped. It may still complete in the background. Check the gallery before starting the same paid action again.
Prompt and selection problems
The app can stop before generation when required information is missing. Depending on the workflow, check that you have:
- Selected an AI character
- Entered a prompt
- Chosen a source image
- Uploaded the required personal photo
- Selected the permitted number of clothing items for Try-On
- Chosen an available voice or model
Read the message shown beside the control and complete only the missing step.
Photo upload problems
Use a supported image format within the file-size limit shown by the app. For selfie and Try-On workflows, use a clear photo with one visible adult person.
If the file cannot be read:
- Open it normally on your device to confirm it is not damaged.
- Export or save it again as a common image format.
- Remove unnecessary metadata or very large dimensions by creating a new copy.
- Upload the new copy.
If the image is blocked, read Content Guidelines.
Prompt rejected or safety check failed
Prompt rejected means the submitted content did not pass the safety rules. Remove the prohibited material before trying again.
Safety check failed means the safety service could not complete its technical check. Wait briefly and retry once without repeatedly changing a safe prompt.
Insufficient access or PORTI
A generation may require an active subscription, enough PORTI, or access to the selected model. The app shows the current requirement before the action begins.
Check Plans and Subscriptions before repeating the action.
Public-character editing locks
Creators cannot regenerate the profile visuals or video-call avatar of a public character. Unpublish the character first, make the changes privately, and submit it for review again.
Users can still create their own eligible media experiences with public companions, but they cannot edit the creator's public character profile.
Video generation problems
Check that the source image loads and the prompt clearly describes movement. Video generation can take longer than an image and may continue after a visible timeout.
If playback fails after completion:
- Reload the gallery
- Check the internet connection
- Try the download control
- Open the downloaded file in a compatible video player
Do not start a duplicate generation only because the preview did not play.
Voice problems
If the voice library does not load or a voice generation fails:
- Confirm that you are signed in
- Check the active subscription or paid-access requirement
- Try the library again after a short wait
- Keep the character page open while a voice is processing
- Confirm that the completed voice is saved and active
Character publication requires an active saved voice.
Contact Support
If the issue continues, include the character, feature, approximate time, exact message, and whether the item later appeared in a gallery. Read Reporting and Contacting Support.