Upload the target video and a portrait reference
Upload the video clip (up to 120 seconds) and a clear frontal portrait photo of the face to swap in. Both inputs are required for the face-replacement workflow.
AI video effects
AI Video Face Swap
Swap faces in any video with AI for realistic, frame-consistent results. Preview the workflow now while the full video runtime rolls out soon.
Previewvideo
Face Swap Video
Preview the video face swap workflow while the dedicated runtime capability is unavailable.
Coming soon
This workflow is listed for planning and will become runnable after model/API support is available.
How it works
How to use Face Swap Video
Preview the face swap workflow
The workflow uses Sora 2, Runway, or PixVerse for face region detection and replacement. Preview the expected swap in the settings panel while the dedicated temporal-identity endpoint is being connected.
Submit for full processing when available
Submit the swap order. Frame-consistent face replacement will be processed once the dedicated video face-swap endpoint is generally available. Current status: preview.
Why use it
Why use AI Pin Maker video tools
Workflow-first setup
Each page maps one user job to the required video inputs and model family.
Model fit
No live endpoint currently supports this exact video workflow, so generation is disabled.
Preview roadmap
Preview tools make upcoming video workflows visible before launch.
Connected routes
Related video tools help continue from effects into editing, music, and production workflows.
FAQ
Video tool FAQ
- Is AI video face swap available now?
- The workflow is in preview mode — the upload flow and settings are live, but full frame-consistent face swapping launches once the dedicated video-replacement endpoint is connected.
- Why isn't video face swap live yet?
- Current AI video generation models can alter appearance per-frame, but maintaining consistent identity across all frames of a video clip requires a dedicated temporal-identity model not yet available on this endpoint.
- What reference photo gives the best face swap results?
- A frontal, well-lit portrait without glasses or occlusion gives the best identity alignment. Side-angle references reduce accuracy because the model can't fully reconstruct the 3D face geometry from a partial view.
- Is there a still-image face swap available right now?
- Yes — the AI image tools section includes ready-to-use face swap for photos. For video face swap, the preview mode lets you queue a job and you will be notified when the video endpoint launches.
Related tools
Explore more video tools
Image to Videovideo
Video effectsReady
AI Kissing Video Generator
Create romantic AI kissing videos from photos or text. Choose type presets for natural couple motion, perfect for sweet, shareable short clips.
kissingcoupleromance
ImageOpen tool
Video Editvideo
Creation toolsReady
AI Video Editor
Edit short videos with AI using prompts and reference images. Restyle, adjust, and refine clips through the verified video edit route, no timeline needed.
video-editorvideo-editreference-images
Video and reference imagesOpen tool
Video Editvideo
Creation toolsReady
AI Video to Video Generator
Convert and restyle uploaded videos with AI. Apply new visual styles while keeping core motion and scene continuity through the verified edit endpoint.
video-editstyle-transfervideo-to-video
VideoOpen tool