Make a photo or video talk in Arabic (lip-sync)

Goal

Create a short talking-head clip in Arabic from a photo (or an existing video) and a script, with matching mouth movement.

Steps

1. Write the script (under 45 seconds of speech). Add tashkeel to names or words the voice might misread.
2. Generate the Arabic voice in ElevenLabs (text to speech), pick a voice that suits the dialect, and download the MP3. Or record your own voice.
3. Upload a front-facing, well-lit photo (or a short video of the face) to HeyGen or Hedra and choose the photo-avatar / audio-to-video option.
4. Upload the MP3 as the audio track and generate. Review mouth sync on 'م ب ف' sounds; regenerate if it drifts.
5. Add captions and export.

Prompt

Replace anything in { } with your details.

Rewrite this script for a {DIALECT} Arabic talking-head video of about {SECONDS} seconds. Keep sentences short (max 12 words), natural and spoken, not formal MSA unless I ask. Add tashkeel only on words a text-to-speech voice could mispronounce. Mark pauses with "…".
Script: {SCRIPT}

Expected output

A captioned talking-photo or lip-synced video in Arabic, typically 20–45 seconds.

Gotchas

• ⚠ Only use faces and voices you own or have written consent for; most tools require it and impersonation can be illegal.
• ⚠ Gulf dialect quality varies by voice; test 2–3 voices.
• ⚠ Side profiles, glasses glare and beards reduce sync quality.
• ⚠ Arabic quality of each tool is unverified until our test.