mcpField
F
fal-lip-sync

fal-lip-sync

github.com/fal-ai-communityCreative & media258104 installsUpdated 24w ago
Repository

About fal-lip-sync

fal-lip-sync is a skill for generating talking-head videos from portraits and audio or text, and for syncing new audio to existing video. It is for workflows that need model-driven lip synchronization and portrait animation through fal.ai.

Key features

  1. 01Generate talking-head video from image and audio
  2. 02Create speech from text with TTS
  3. 03Sync new audio to existing video
  4. 04Choose model endpoint and run mode

Use cases

  1. 01Animate a portrait to deliver spoken text
  2. 02Make a video match a replacement audio track

MCP support

The skill references the search_models MCP tool to find lip-sync and talking-head models.

Install

Copy the command and run it in a terminal.

$npx skills add https://github.com/fal-ai-community/skills --skill fal-lip-sync

After installing, type/fal-lip-syncin Claude Code to confirm it loaded.

Bundled filesscripts/lip-sync.shscripts/talking-head.sh

Details

From the project README

fal-lip-sync

Create talking head videos, sync lips to audio, and animate portraits using fal.ai models.

Scripts

ScriptPurpose
talking-head.shGenerate a talking head video from an image + audio/text
lip-sync.shSync lips in an existing video to new audio

Usage

Talking Head (Image + Audio → Video)
./scripts/talking-head.sh --image-url "https://example.com/portrait.jpg" --audio-url "https://example.com/speech.mp3" --model veed/fabric-1.0
Talking Head (Image + Text → Video with auto TTS)
./scripts/talking-head.sh --image-url "https://example.com/portrait.jpg" --text "Hello, welcome to our presentation" --model fal-ai/creatify/aurora
Lip Sync (Video + Audio → Synced Video)
./scripts/lip-sync.sh --video-url "https://example.com/video.mp4" --audio-url "https://example.com/new-speech.mp3"

Arguments

talking-head.sh
ArgumentDescriptionRequired
--image-urlURL of portrait/face imageYes
--audio-urlURL of audio to syncYes (or --text)
--textText to speak (auto TTS)Yes (or --audio-url)
--model / -mModel endpointNo (default: veed/fabric-1.0)
--tts-modelTTS model for --text modeNo (default: fal-ai/minimax/speech-2.6-turbo)
--wait / -wWait for completionNo (default: true)
--async / -aReturn request ID immediatelyNo
lip-sync.sh
ArgumentDescriptionRequired
--video-urlURL of video to lip syncYes
--audio-urlURL of audio to sync toYes
--model / -mModel endpointNo (default: fal-ai/sync-lipsync/v2)

Finding Models

To discover the best and latest lip sync and talking head models, use the search API:

# Search for talking head models
bash /mnt/skills/user/fal-generate/scripts/search-models.sh --query "talking head"

# Search for lip sync models
bash /mnt/skills/user/fal-generate/scripts/search-models.sh --query "lip sync"

# Search for live portrait / expression transfer
bash /mnt/skills/user/fal-generate/scripts/search-models.sh --query "live portrait"

Or use the search_models MCP tool with relevant keywords like "lip sync", "talking head", "avatar".

Output Format

{
  "video": {
    "url": "https://fal.media/files/...",
    "content_type": "video/mp4"
  }
}

Present the video URL directly to the user.

Comments

Markdown supported