Follow every control from account sign-in and source upload through extension direction, output settings, generation status, final review, and download.
AI Video Extender is an authenticated workflow. Sign in before selecting a file so the upload and generated task can be linked to your account. New accounts receive the credit balance shown in the header, and additional credits can be reviewed on the Pricing page.
Prepare one MP4, MOV, or WebM source no larger than 100 MB. Use the version with the exact opening or ending you want to extend. The tool uploads the source, extracts the selected boundary frame, and sends that reference with the generation request, so choosing the correct edit matters.
This page explains the controls and status messages in the current interface. For creative prompt formulas, use the AI Video Extender Prompt Guide. For quality-preservation methods and editing alternatives, read How to Make a Video Longer Without Losing Quality.
Complete the workflow in order. Each stage prepares information required by the next, and the interface displays a status message when an input is missing or a background operation is still running.
Open AI Video Extender and sign in from the header or by selecting the upload area. Uploading and generation require an authenticated account because the service uses your account to store tasks, display available credits, and return completed work in My Creations. After sign-in, wait briefly for account authorization to finish before selecting the source file.
If the account button still shows a retry state, complete authorization before uploading media.
Select Upload a video and choose an MP4, MOV, or WebM file up to 100 MB. The interface uploads the complete source and prepares the boundary frame used for generation. Keep the page open until progress reaches completion and the frame preview appears. Verify that the filename, source scene, and visible preview match the clip you intended to use.
Do not press Generate while the upload status says the file or boundary frame is still being prepared.
Choose Extend after video when the generated footage should follow the original. The tool extracts the last frame and orders the finished sequence as source first, generated section second. Choose Extend before video when you need an earlier lead-in. The tool then extracts the first frame and places the generated section before the uploaded source.
Changing direction after upload automatically prepares the correct first or last frame again, so wait for the new preview.
Enter one clear instruction in What should happen next. State the subject action and, when relevant, the camera behavior. For a forward extension, begin where the source ends; for a backward extension, describe an earlier moment that can naturally arrive at the opening frame. The field allows detailed instructions, but a focused request is usually easier to evaluate.
Example: The cyclist continues along the coastal road while the camera tracks steadily from behind.
Select 16:9 Landscape, 9:16 Portrait, or 1:1 Square according to the destination of the finished video. Choose 768p for faster concept testing or 2K when the project needs more detail. Set the generated section to 5 or 10 seconds. These controls describe the requested output; they do not crop or repair the uploaded source before generation.
Match the source orientation whenever possible to avoid introducing an unnecessary composition change at the join.
Review the prepared frame, prompt, direction, duration, ratio, and resolution, then select Extend Video with AI. The button validates that you are signed in, the source is ready, and a prompt is present. A successful request either returns a result directly or creates a background task. Long-running work remains associated with your account while processing continues.
Do not submit several identical tasks because the first click appears slow; read the status message before trying again.
Open My Creations to follow processing, ready, and failed tasks. When a result is ready, watch the complete sequence rather than judging only the generated segment. Confirm the order is correct for the selected direction and check the transition between source and extension. Download the complete video only after it serves the intended edit.
Use Refresh when a task has finished remotely but its status has not yet updated in the current view.
Configure the request after the source and direction are ready. The most suitable choice depends on where the result will be used, how much motion the next beat contains, and the level of detail required for review.
Extend after uses the source ending as the new start. Extend before creates a lead-in that arrives at the source opening.
Choose 16:9 for landscape, 9:16 for vertical publishing, or 1:1 for square placements.
Use 768p for economical iteration and 2K when the output needs additional visual detail.
Select 5 seconds for one concise beat or 10 seconds when the action needs more room to develop.
The prompt field tells the service what should change after or before the selected boundary. A concise subject action plus camera behavior is enough to begin. Avoid turning this step into a complete screenplay: the uploaded frame already communicates the visible character, object, setting, and composition.
“The cyclist continues along the coastal road while the camera tracks steadily from behind.”
Describe what happens immediately after the source ending.
Describe an earlier movement that naturally arrives at the opening frame.
Status messages explain whether the interface is waiting for a local input, uploading media, running generation, or displaying a completed task. Read the current state before repeating an action.
The source video or extracted boundary frame is still transferring. Keep the page open and wait for the ready message.
The request was accepted and new footage is being created. The preview updates when a playable result becomes available.
The task has a result that can be reviewed. Watch the full source-plus-extension sequence before downloading or publishing.
The task could not complete. Read the returned message, verify credits and inputs, then revise only the cause before retrying.
Start with the message displayed by the tool. Most blocked actions relate to authentication, unsupported media, an unfinished upload, a missing prompt, task processing, or account credits.
This is expected for signed-out visitors. Finish sign-in, allow account authorization to complete, and select the upload area again.
Confirm the file is MP4, MOV, or WebM, is no larger than 100 MB, and that account authorization is ready.
Check whether Extend after or Extend before is selected. Changing direction makes the tool extract and upload the corresponding frame again.
The prompt field cannot be empty. Add a direct instruction describing the immediate action or shot that should be generated.
Generation can continue in the background. Open My Creations, use Refresh, and avoid creating duplicate tasks while the first one is active.
Review the on-screen message, available credits, upload readiness, and request settings. Correct the identified issue before submitting again.
Open the ready item in My Creations and watch through the boundary at normal speed. The generated section should not be evaluated in isolation because usefulness depends on how it joins the uploaded source and whether the selected sequence order is correct.
Source then generated for Extend after; generated then source for Extend before.
The boundary should connect without an obvious jump in position, motion, lighting, or framing.
The main person, object, vehicle, or product should remain recognizable across the join.
The new camera path should agree with the movement established near the source boundary.
Confirm duration, aspect ratio, resolution, and whether the complete video is ready for the target platform.
Review generated content and confirm you have permission to use the uploaded source and resulting media.
Yes. Upload and generation require sign-in so tasks, credits, and completed videos can be associated with your account.
The current upload control accepts MP4, MOV, and WebM files with a maximum file size of 100 MB.
Extend after generates footage following the source ending. Extend before creates footage leading into the source opening and changes the final sequence order accordingly.
The tool uses the last frame for a forward extension and the first frame for a backward extension, so it prepares a different boundary reference.
You can request 16:9 Landscape, 9:16 Portrait, or 1:1 Square output from the current workspace.
Yes. The resolution selector offers 768p and 2K. Higher-resolution generation uses the credit cost shown by the interface.
Yes. Use the duration control to select either 5 or 10 seconds before submitting the task.
Open My Creations while signed in. It lists processing, ready, and failed work associated with your account.
The extension workflow retains the source association and presents the original plus generated footage in the order selected by the extension direction.
Use the dedicated Prompt Guide for formulas, examples, templates, negative instructions, and detailed prompt troubleshooting.
Sign in, prepare the correct boundary frame, choose the output settings, submit one focused continuation, and review the finished sequence in My Creations.