n8n Workflow: Multilingual Video Dubbing with Claude + Shotstack

Turn one video into multilingual dubbed copies—with captions, timing, and voices on point

This n8n workflow takes a single video link and a list of languages, then produces dubbed and captioned versions for each language. Using Shotstack for timed transcription and rendering, and Claude for translation, it schedules an ElevenLabs voice to speak each translated line exactly when the original line occurs. Built for self-hosted n8n, it saves every rendered output to disk.

What this workflow does

• Transcribes the video into timed captions using Shotstack’s Ingest API.
• Translates each caption per target language with Claude—keeping translations short enough to fit the original time window.
• Verifies translation timing: a Code step sends any mismatch back to Claude for correction.
• Renders one video per language in Shotstack, lowering the original audio and adding the ElevenLabs voice track with the translated lines synchronized to the captions.
• Outputs captioned dubbing in the rendered video, and saves each version into a “translated” folder named after the video and language.

Use cases

• SaaS teams localizing product explainers and marketing videos into multiple languages.
• Automation engineers building a reusable multilingual video dubbing pipeline in n8n.
• Content operators generating consistent dubbed + captioned assets from one source video.

Technical details

• Shotstack (Transcribe + Check transcript + Make the video + Check the render) with a production API key.
• Claude (Anthropic) for translating timed captions, with a Code step to handle timing mismatches.
• ElevenLabs voice generation per line, synchronized to the original speech timing.
• n8n self-hosted file output: creates and writes to an empty translated folder (e.g., Docker path /home/node/.n8n-files), saving renders on disk.
• Workflow nodes used include: if, set, code, no op, wait, and sticky note.

For setup, install the Shotstack community node, add your Claude and Shotstack API keys, configure the translated output folder, then provide your video link and languages (comma-separated) to execute.