sync/sync

MCP Server

Sync.so MCP Server - lipsync postprocessing for video

Sync.so (Sync Labs) MCP server. State-of-the-art lipsync. Takes a source video + a target audio track and produces a perfectly lipsynced output. Drop-in postprocess after voiceover replacement, dubbing, or AI-avatar generation.

MCP Server

  • Package: @sync.so/mcp-server (community)
  • Transport: stdio
  • Command: npx -y @sync.so/mcp-server

Authentication

VariableRequiredDescription
SYNC_API_KEYYesAPI key from sync.so dashboard

Available Tools

ToolDescription
generate_create-generationCreate a lipsync generation
generate_get-generation / generations_get-by-idPoll generation status
generate_estimate-costEstimate generation cost
assets_* (create, get, update, delete, create-upload-url)Manage media assets
batch_* / projects_*Batch jobs and project organization
tts_create / voices_*TTS and voice management

Configuration Example

toml
[[mcp.servers]]
name = "sync"
type = "stdio"
command = "npx"
args = ["-y", "@sync.so/mcp-server"]
timeout_seconds = 120
env = { SYNC_API_KEY = "{{INPUT:SYNC_API_KEY}}" }
tools = []

Notes: Pricing is per-second of output (~$0.10/s). Use only on the final selected cut, not on every variant.