Skip to main content
VEED Dubbing translates the speech in a clip into another language. Every speaker’s voice is cloned and their lips are re-synced to the translation, so the result looks spoken in the target language instead of captioned. In ComfyUI the node is VEED Dubbing, with two model tiers. standard is the cheaper one: mono audio, 85 target languages, with Spanish as Latin American and Portuguese as Brazilian. pro adds regional accents and stereo audio, and reaches 101 target languages, including variants such as English (UK), Spanish (Mexico) and Chinese (Taiwan). It is a cloud API node. It needs a Comfy account with credits, downloads no model files, and bills per second of output. See Partner node pricing for current rates.
To use the Partner Nodes, you need to ensure that you are logged in properly and using a permitted network environment. Please refer to the Partner Nodes Overview section of the documentation to understand the specific requirements for using the Partner Nodes.
Make sure your ComfyUI is updated.Workflows in this guide can be found in the Workflow Templates. If you can’t find them in the template, your ComfyUI may be outdated.If nodes are missing when loading a workflow, possible reasons:
  1. You are not using the latest ComfyUI version (Nightly version)
  2. Some nodes failed to import at startup

​
Available workflow

​
VEED Dubbing: video translation

Translate a clip you already have. The performance, the shot and the timing stay as recorded; what changes is the language the speaker is heard in. VEED Dubbing workflow preview

Download Workflow

Download JSON or search “VEED Dubbing: Video Translation” in Template Library
Input materials

api_veed_dubbing_input_video.mp4

Load into the LoadVideo node. One speaker facing the camera
The clip runs up to 599 seconds and the output is as long as its audio track, at 25 fps.

​
Node controls

​
Tips

  • Send footage with a visible speaker. Dubbing works on what it can see, so a single person facing the camera is the reliable case. Backs of heads, hands, product shots and crowd scenes do not dub.
  • Check that the audio actually has speech. Narration, dialogue and interviews all work. Music, sound effects and silence give the model nothing to translate.
  • Name the source language when auto mishears. auto handles clean recordings, but a short clip or a strong accent dubs more accurately when you set source_language yourself.
  • Pick pro when the accent matters. The standard tier gives one variant per language, so requests for British English or Mexican Spanish only exist on pro.
  • Cut to the length limit first. The node accepts up to 599 seconds of video, and the output is as long as the audio track, so trim before the run rather than after.

​
Get started

  1. Update ComfyUI to the latest version
  2. Go to Template Library, search for VEED Dubbing
  3. Load a talking-head clip into the LoadVideo node
  4. Choose the tier in model, set target_language, then run the workflow