Video Face Swap API

What is a Video Face Swap API?

A video face swap API takes a source photo of a face and a target video, then replaces the face throughout every frame of that clip — preserving expression, lighting, and frame-by-frame consistency so the result reads as a natural clip rather than a shaky composite.

Muapi exposes this as a single endpoint, ai-video-face-swap: one API key, one request/poll flow, priced by video length (roughly $0.10/sec), no moderation filter applied to the output. Only need to swap a face in a single photo instead of a video? See the Image Face Swap API.

Key Capabilities

Input Requirements

• Target video must be under 10MB, resolution below 1280×720 (or 720×1280 portrait), and no more than 600 frames.
• Use a clear, front-facing source photo. Good lighting and an unobstructed face give the model the cleanest reference for blending.
• Expect some iteration on complex scenes. Extreme angles, heavy occlusion, or very small faces in the target reduce swap quality — test before batch-processing.

Other Ways to Swap a Face in Video on Muapi

ai-video-face-swap is the dedicated, cheapest option for a straight face-only swap. A couple of motion-transfer models can also replace who appears in a clip, at a higher cost, while additionally re-animating the character's full body to match the reference performance:

Kling copies motion from a reference performance onto a new character image; Wan 2.2 Animate replaces the subject in an existing real scene while preserving the original camera work and background — closer to a true swap-in-place. Want the whole performer recast instead of just motion-transferred? See Seedance Motion Control. Need the same face consistent across brand-new generated video, not swapped into an existing clip? See the AI Character Consistency API.

Example Prompt for Motion-Transfer Editors

On Kling Motion Control or Wan 2.2 Animate, pass the new character image and the reference video, with a prompt like:

"Apply the motion, expression, and camera work from the reference video to this character image. Keep the original environment and background unchanged, and preserve the performance's timing exactly."

Consent and Responsible Use

Video face swap generates a likeness-based clip of whoever appears in the source photo. Only submit photos of people who have agreed to be used this way — yourself, a hired model under a release, or a customer who explicitly opted in to a face-swap feature. Do not use someone else's likeness without their permission, and do not use this endpoint to impersonate a real person without consent. Muapi does not verify consent on your behalf — that responsibility sits with the API caller, consistent with the platform's general terms of use.

How to Call the Video Face Swap API

1. Host your image and video. Upload via POST /api/v1/upload_file, or use any publicly reachable URL.
2. Submit the request. POST /api/v1/ai-video-face-swap.
3. Poll for completion. Check GET /api/v1/predictions/{request_id}/result until status is completed, then download the output.

curl -X POST https://api.muapi.ai/api/v1/ai-video-face-swap \
  -H "Content-Type: application/json" \
  -H "x-api-key: YOUR_API_KEY" \
  -d '{
    "image_url": "https://example.com/source-face.jpg",
    "video_url": "https://example.com/target-clip.mp4"
  }'

import requests

response = requests.post(
    "https://api.muapi.ai/api/v1/ai-video-face-swap",
    headers={"x-api-key": "YOUR_API_KEY"},
    json={
        "image_url": "https://example.com/source-face.jpg",
        "video_url": "https://example.com/target-clip.mp4",
    },
)
request_id = response.json()["request_id"]

result = requests.get(
    f"https://api.muapi.ai/api/v1/predictions/{request_id}/result",
    headers={"x-api-key": "YOUR_API_KEY"},
)
print(result.json())

Frequently Asked Questions