VideoDubber Documentation

Welcome to VideoDubber

VideoDubber.ai is an AI-first video translation and dubbing platform built for YouTubers, marketers, educators, filmmakers and businesses. Upload a video (or paste a YouTube link), pick a target language, and get a natural-sounding dubbed video back — with synchronized subtitles and, optionally, your own cloned voice.

Quickstart: Your First Dub in 4 Steps

The fastest path from source video to a polished, localized dub. You can start a free trial on any plan to test the full workflow before being charged.

Create an Account

You can sign up in seconds and start a free trial on any plan from inside the dashboard. VideoDubber supports sign-in with Google and with email/password.

1. Go to videodubber.ai/login.
2. Choose Continue with Google or create a password account.
3. Verify your email (only required for password accounts).
4. You’ll land on the dashboard where you can start a free trial on any plan and begin your first project.

Supported Formats & Limits

Input formats

• Video: MP4, MOV, WebM, MKV, AVI, FLV
• Audio: MP3, WAV, M4A, AAC, FLAC, OGG
• Subtitles: SRT, VTT (for subtitle-only translation)
• URLs: Public YouTube links (and select other platforms)

Output formats

• Video: MP4 (H.264), with optional burned-in subtitles
• Subtitles: SRT and VTT in any translated language
• Audio-only: MP3 and WAV exports of the dubbed track

Video length and size

AI Video Translation

Translate a source video into any of 150+ languages while preserving tone, pacing, and emotion. Our pipeline separates the speaker’s voice from the background, transcribes the audio, translates the text with a context-aware model, and re-synthesizes speech in the target language.

How the pipeline works

• Vocal isolation:background music and ambient sound are separated from the speaker’s voice before transcription.
• Speech-to-text: a multilingual ASR model produces a time-aligned transcript with punctuation and speaker labels.
• Translation: the transcript is translated by a large language model tuned for spoken conversational style, not literal text.
• Text-to-speech: the translated text is spoken by the selected AI voice (or your cloned voice) with matching prosody.
• Re-mixing: the new voice track is mixed back with the original background audio so ambience is preserved.

Voice Cloning

Voice cloning reproduces your own vocal identity so your dubbed videos sound like you in every language. This is especially valuable for personal-brand creators, executives, and educators who rely on a recognizable voice.

What you need

• A clean voice sample, ideally 30–90 seconds in a quiet environment, with a single speaker (you).
• Natural, conversational speech — not a scripted monotone.
• No background music, reverb, or echo.
• Any paid plan — instant voice cloning is included from Starter up, and premium voice cloning unlocks on Pro and above.

Languages supported for cloning

Your cloned voice can speak any of the 150+ languages in VideoDubber, even if you only provided samples in one language. The model transfers your timbre while producing native-sounding pronunciation in the target language.

AI Lip-Sync

Lip-sync modifies the speaker’s mouth region on a frame-by-frame basis so it matches the dubbed audio. This makes talking-head and to-camera videos look natural in the target language rather than obviously dubbed.

When to use it

• Talking-head YouTube videos, interviews, and tutorials.
• Customer testimonials and sales videos.
• Educational and e-learning content with a visible instructor.

When to skip it

• Voice-over videos where the speaker is off-screen.
• Heavily animated or cartoon content.
• Videos with frequent face occlusion (hands, masks, microphones).

Subtitle Editor

Every project includes a browser-based subtitle editor where you can correct transcripts, fine-tune timing, split or merge lines, and apply a custom dictionary for brand names and technical terms.

What you can do

• Edit the source transcript and the translation side-by-side.
• Shift timing per line, or nudge the entire track by a fixed offset.
• Maintain a brand dictionary so product names, acronyms and proper nouns are preserved verbatim across all target languages.
• Regenerate audio for a single line after editing — you don’t need to re-render the whole video.
• Export subtitles as SRT or VTT, or burn them into the video at export.

1
00:00:01,000 --> 00:00:04,500
Welcome to the future of video localization.

2
00:00:04,800 --> 00:00:08,200
Translate, dub, and publish in minutes.

Text-to-Speech

Paste text (or upload a script) and generate a natural-sounding narration in any supported language and voice. Useful for voice-overs, podcast intros, audiobooks, e-learning modules, and accessibility tracks.

Controls available

Multi-Speaker Detection

VideoDubber automatically diarizes your audio — detecting when each distinct speaker is talking — and lets you assign a different voice to each one. This is essential for podcasts, interviews, panel discussions, and fiction.

• Up to 10 speakers detected per project.
• Assign any voice from the library (or a cloned voice) per speaker.
• Edit speaker labels if diarization merges or splits people incorrectly.
• Works alongside lip-sync: each detected face gets the correct voice.

Background Music Controls

When we dub a video we separate the speaker’s voice from the background audio. You decide what happens to that background stem:

• Keep original: preserve the existing music and ambience under your new voice track (most common choice).
• Silence it: mute the background completely and output clean dubbed speech on silence.
• Replace: pick a track from our royalty-free library or upload your own .mp3/.wav to use as the new bed.
• Ducking: automatic volume ducking so music drops whenever the speaker is talking and rises during pauses.

Supported Languages & Accents

We support over 150 languages and regional accents, covering roughly 99.86% of the world’s internet-connected native speakers. Highlights include:

For a full, searchable list including dialect codes, head to the language picker inside any project — the menu is always up-to-date as we add new voices.

Workflow: Dub a YouTube Video

The fastest way to localize existing YouTube content for new language markets.

Workflow: Localize an Entire Course

Recommended approach for e-learning creators producing multi-language courses at scale.

1. Build a brand dictionaryfirst — add all product names, proper nouns, and technical terms. This ensures consistency across lessons.
2. Clone the instructor’s voice on any paid plan so every lesson in every language sounds like the same teacher.
3. Upload lessons in batches and assign the same target languages to all of them so settings are reused.
4. Once generated, spot-check every lesson in the editor and use Regenerate line for any sentence you rephrase.
5. Export burned-in subtitles for platforms that don’t support SRT, and soft SRT for LMS platforms that do.

Workflow: Translate Subtitles Only

If you only need captions in additional languages — no new voice track — VideoDubber can translate and re-time an existing subtitle file.

1. Upload an existing .srt or .vtt file, or point VideoDubber at a YouTube URL and use the auto-generated transcript.
2. Select one or more target languages.
3. The translated subtitle files are produced in seconds and can be downloaded in SRT or VTT format.

Workflow: Localize a Podcast

VideoDubber handles audio-only projects with the same pipeline as video — just skip lip-sync.

1. Upload your episode’s .mp3, .wav, or.m4a file.
2. If you have two or more hosts, turn on Multi-speaker detectionand assign a unique voice per host.
3. Pick one or more target languages.
4. Export translated audio as MP3 or WAV and publish to your feed as a separate localized show.

Using the Brand Dictionary

A brand dictionary is a per-account glossary that tells VideoDubber how to handle words that should not be translated (product names), custom pronunciations (acronyms), and preferred translations for ambiguous terms.

Entry types

• Do-not-translate: keep the term verbatim in every language. Example: VideoDubber, YouTube.
• Phonetic override: spell out how an acronym should be spoken. Example: SaaS → "sass".
• Preferred translation: lock in a specific translation for a term that otherwise has multiple valid options.

Export Options

At export time you can customize the final deliverable:

• Video container: MP4 with H.264 video and AAC audio.
• Resolution: 480p, 720p, 1080p, 2K, 4K (subject to source).
• Subtitles: burn into the frame or keep as a separate SRT/VTT.
• Audio track:include original + dubbed as multi-track audio (useful for YouTube’s multi-language audio feature).
• Standalone audio: MP3 (lossy) or WAV (lossless) dubbed audio only.

Plans Overview

VideoDubber offers four paid plans plus a free trial on every plan, so you can test the product end-to-end before committing. All plans include voice cloning, multi-speaker studio, no watermark, unlimited video length, and access to premium translation models. You can upgrade, downgrade, or cancel any time from billing settings.

Plan minutes and pricing are displayed per region; Indian accounts are automatically shown equivalent INR pricing. Full, always-up-to-date plan comparison lives on the pricing page. Every plan ships with a free trial — start one with the Start Free Trial button on any plan card.

How Minutes Are Counted

VideoDubber meters usage by the duration of the source audio/video, not the output. Each plan includes a monthly pool of translation minutes, and you can keep working beyond the pool with per-minute extra usage.

Payments & Invoicing

Accepted payment methods

• All major credit and debit cards (Visa, Mastercard, Amex, Discover).
• Apple Pay and Google Pay where available.
• UPI, net banking, and local cards for Indian accounts.
• Wire transfer and PO-based invoicing on annual Scale contracts.

Billing cycle

• All paid plans renew automatically every month (or every year on annual billing).
• Unused minutes do not roll over into the next cycle.
• You can upgrade mid-cycle and pay a prorated difference; downgrades take effect at the next renewal.
• Invoices are emailed automatically and available underSettings → Billing.

Refunds & Cancellation

You can cancel your subscription anytime from Settings → Billing. Cancellation takes effect at the end of the current billing period — you keep full access until then and your projects remain accessible.

Refund policy

• If VideoDubber fails to deliver a usable dub because of a platform error, we refund the minutes and, where appropriate, the charge.
• We do not offer refunds for minutes consumed on successful jobs that were later abandoned or replaced.
• Annual Scale contracts have custom refund terms defined in your agreement.

For refund requests, email contact@videodubber.ai with your account email and a brief description of the issue.

Data Privacy & Security

• All uploads are transferred over HTTPS and stored encrypted at rest.
• Your source videos are never used to train public models.
• Voice samples for cloning are stored encrypted and tied only to your account. You can delete any cloned voice at any time.
• Account deletion wipes your projects, uploads, and voice clones within 30 days.

Read the full privacy policy and terms of service.

Team & Multi-User Access

Team seats and shared workspaces are available on the Scale plan. A workspace lets multiple editors access the same project library, share brand dictionaries, and collaborate on long-running localization programs.

1. Contact contact@videodubber.ai with your expected team size and use case.
2. We configure a workspace, transfer your existing projects, and invite your teammates by email.
3. Roles include Admin (billing + projects), Editor (create + edit), and Viewer (read-only).

Upload & Import Issues

My upload is stuck or fails

• Check your file is within the upload size limit for your plan.
• Make sure the container is one of MP4, MOV, WebM, MKV, AVI, or FLV.
• Try again on a stable connection — browser uploads are resumable, but extreme packet loss can still fail.
• If your source is larger than 2 GB, export a 1080p MP4 first using any video editor and upload that instead.

YouTube URL won’t load

• The video must be public or unlisted. Private and age-gated videos cannot be fetched.
• Livestreams must be finished and available on-demand before import.
• Regional blocks can stop us from fetching certain videos; in that case, download the file yourself and upload it directly.

Fixing Translation & Voice Quality

Most quality issues trace back to source audio. A clean source input is the single biggest lever you have.

If the translation is off

1. Open the editor and correct the source transcript first — translation errors usually come from ASR errors upstream.
2. Add proper nouns and jargon to your brand dictionary.
3. Regenerate only the affected lines using Regenerate line.

If the voice sounds robotic or rushed

• Try a different voice in the same language — tone varies a lot between voices.
• Adjust speed slightly (0.95x often sounds more natural than 1.0x for long sentences).
• Shorten overly long lines with line-splits in the editor to give the TTS more breathing room.

If voice cloning sounds wrong

• Re-record your voice sample in a quieter room with a better microphone.
• Provide 60–90 seconds of natural speech rather than 15 seconds of scripted reading.
• Avoid samples with reverb, background music, or phone compression.

Lip-Sync Doesn’t Look Right

• Face too small: lip-sync works best when the face occupies at least ~15% of the frame. Zoom in or re-crop.
• Fast head motion: rapid turns can cause the model to lose track of the mouth region. Consider disabling lip-sync on action-heavy clips.
• Occluded mouth: hands, masks, microphones or heavy beards can confuse detection.
• Multiple speakers: make sure each speaker has been labeled correctly in the diarization panel.

Export & Download Problems

• Export button greyed out:make sure processing has finished — there is a progress bar at the top of the editor.
• SRT timing off: use the Shift timing tool in the editor to offset the whole track, or regenerate after editing the transcript.
• Download link expired: links are valid for 24 hours; regenerate from the project page any time.

Usage Limit & Error Messages

• QUOTA_EXCEEDED— You’ve consumed your monthly minutes. Upgrade or buy top-up minutes to continue.
• UNSUPPORTED_LANGUAGE— The detected source language isn’t supported for the selected voice. Pick another voice or change the target language.
• AUDIO_TOO_NOISY— Source audio SNR is too low for reliable transcription. Clean up the audio in your editor first.
• FACE_NOT_DETECTED— Lip-sync couldn’t find a face to operate on. Turn off lip-sync or provide a clip with a visible speaker.

If you see any other error code, email supportwith the project ID and we’ll investigate.

Frequently Asked Questions

Short answers to the questions we hear most often. For anything not covered here, reach out at contact@videodubber.ai.