Get the right result: check these first
Actor, product, and words are separate. An actor photo is required; a product photo is optional. Put spoken words in Script and visual directions in Action prompt.
Use your own recording. Select LTX-2.5, then Upload script audio in the Script section. Upload the recording and check its selected section. This upload option is model-specific. LTX-2.5 offers Upload script audio; follow the controls currently shown in Script.
Model or duration changed. Recheck the duration, resolution, and credit estimate after changing models. A longer script does not override the selected clip duration.
Still stuck? Contact support with the tool name, project link or visible ID if available, actor/product inputs, model, script or uploaded-audio selection, duration, and timestamp of the problem. Keep the project while it is reviewed. How to send useful troubleshooting details.
Access
Open AI UGC Ads editor and sign in. Generation uses credits. Available models, resolution and duration can depend on your account and the selected options; follow the access prompt and estimate shown in the editor.
Start with an actor
Select Upload Actor to use your own actor image, or choose a preset AI actor. Select Upload Product if you want to add an optional product image. You do not need to create a saved Character first.
Supported images
Uploads support PNG, JPG, JPEG, JFIF, HEIC, WEBP, AVIF, JP2, TIFF, TIF, and BMP. HEIC is converted for the tool.
Images cannot exceed 8,000px on either side. Upload thresholds are 200 MB on Basic, 2 GB on Creator, 5 GB on Pro, and 10 GB on Business.
Prepare your ad
Upload or choose an actor.
Optionally upload a product.
Enter the words the presenter should say in Script. To supply your own recording instead, select LTX-2.5 and choose Upload script audio, then upload and check the selected audio section.
Optionally use Action prompt to describe the scene or action.
Choose the model shown in the editor.
Choose duration, aspect ratio and resolution from the available Settings.
The current single-page editor does not require the older Generate Variations sequence. Follow the controls you actually see; model choices can change the available settings.
Uploaded audio and video length
After a fresh script-audio upload, duration follows the selected audio section, fitted to the model’s available duration options, until you choose a duration yourself. With uploaded audio, the rendered video and its credit estimate use the shorter of the chosen duration and the selected audio section. A longer duration selection does not extend a shorter recording; a shorter selection can cut off speech. Check the audio trim, displayed output length, and credit estimate before generating. A typed script is different: shorten it or choose a supported longer clip if the speech estimate exceeds the video length.
Create the video
Check your actor, optional product image, and either the spoken script or uploaded script audio.
Review the displayed duration, resolution and credit estimate.
Select Generate Video when ready. Do not assume an older duration list applies to every model.
Review the completed output in History or My Library and download the version you want to keep.
If Upload Actor is missing
Open the AI UGC Ads editor and look for Upload Actor under Image Uploads. If you only see a crop or opening-shot screen, send support a screenshot and the page URL so we can identify the step you are on. Include your browser and account email; do not keep restarting the workflow or purchasing upgrades to look for the control.
Upload errors
Invalid image: convert to JPG or PNG.
Dimensions too large: resize below 8,000px on each side.
File-size message: use a smaller file or follow the plan prompt.
Generation remains in progress: refresh and check History or My Library. Avoid repeated retries; send support the project link or ID, time and visible status if it remains stuck.