Pexo

THE PEXO PROMPT LIBRARY / LISTING

AI Avatar Video Prompts: Plan a Presenter and Voice

A presenter brief needs a character reference, script, voice direction and review of pronunciation and lip sync. The copy-ready briefs below are original starting points. Adapt them to your own subject and assets; no example is presented as a tested product output.

An original presenter in a daylight studio facing an unbranded camera and microphone, with a neutral backdrop and wardrobe swatches
Illustrative concept image made for this page · not a tested product output.

Avatar Look and Voice Prompts

Give the presenter an authorized character reference, approved script and voice direction. Review face and wardrobe continuity.

What must be fixed before choosing an avatar look?

Keep one authorized character reference, the approved 80-word script and the permitted voice as separate inputs. The visual can change lighting or framing, but it should not quietly change the face, wardrobe or spoken claims. Review the first and last frame together, then replay a long word and a pause with the source audio.

Copyable starter brief

Prompt text
Create a presenter video from the authorized character reference and approved 80-word script. Keep wardrobe, background and face consistent; use a calm delivery and review lip sync before use.

Which brief fits your source?

Brief 1: Authorized presenter and 80-word script. Brief 2: Fictional illustrated multilingual host.

Selection rule: Use the authorized presenter for continuity with a supplied face; use the fictional host when no real identity should be depicted.

Source-to-instruction relationship: A real character reference and approved script fix identity and claims; translation adds a separate pronunciation review.

If the draft misses the brief

Visible signal: The face or wardrobe drifts between frames, or translated delivery changes the approved claim.

Next revision: Freeze the character reference and script, then review the translated line with a fluent speaker before changing voice style.

Keep the source and review together

Source to retain: the authorized character reference, approved script and translation where used. After an actual run: save the exact prompt version, resulting file and a review note stating whether the face and wardrobe hold across frames and the spoken wording remains approved. Until that run exists, leave the result and review fields unfilled. The Gallery cases below are separate published projects and cannot complete this record.

Multilingual Delivery

For translated delivery, use approved translated text and a pronunciation review. Do not imply every language or voice is available in every account.

Copyable starter brief

Prompt text
Design a fictional illustrated host for a short explainer. Specify a blue jacket, neutral studio background and measured voice. Use the approved translation and check pronunciation with a native reviewer.

Separate the avatar brief from the spoken script

Keep the approved words in one field and visual direction in another. In the visual field, list the authorized face reference, wardrobe and background; in the delivery field, mark names, numerals and any pause before the call to action. For a translated script, ask a fluent reviewer to approve the wording before checking lip movement. A good-looking mouth animation does not establish that the translation is correct.

How to Write Your Own AI Avatar Video Prompts

Provide an authorized reference, exact script, wardrobe and voice direction. For translated speech, bring approved wording and check names with a fluent reviewer. Do not assume every avatar or language option is available.

Four fields for the presenter brief

Keep approved script as the exact spoken words; put face, wardrobe and background in a separate visual field. Use delivery for pace and pauses, then record names and numerals in a pronunciation review field. For the illustrated host, the blue jacket and neutral studio are visual constraints, not speech instructions. Leave any unknown script field blank until its owner approves it.

Copyable presenter handoff fields

Fill these fields before drafting the final video brief; leave the script blank until its owner approves the exact words.

  • Authorized character reference: [file or character sheet and permission status]
  • Visual continuity: [face, wardrobe, background and fixed details]
  • Approved spoken script: [paste exact approved wording; otherwise leave blank]
  • Delivery and pronunciation review: [pace, pauses, names and numerals to check with the source]

The brackets are a planning template, not a record of a tested avatar or spoken result.

Use a Reviewed Video Brief in Pexo

Bring your brief and approved assets to Pexo's AI avatar workflow. Review the proposed direction and preview, request specific changes in conversation, and check the final video for factual, visual and rights accuracy. Account credits and eligible free use depend on the current plan.

Review the avatar as a continuing character

Use this brief when: Use a fictional host for a new identity; use a real person's likeness or voice only with the needed permission.

Acceptance check: Compare face, clothing and background across the opening, middle and ending, and listen for the approved script and pronunciation.

If it fails: If identity drifts, shorten the scene or return to one approved reference instead of adding more style cues.

Gate the presenter handoff before rendering

The 80-word script in the starter brief is an input the reader must supply; the page has no approved script or recorded presenter output. A handoff is ready only when the owner has filled all four rows below:

Field Required source Reject the handoff when
Character Authorized reference or original character sheet Face, wardrobe or permission status is unspecified.
Speech Exact approved script, including names and numerals A visual direction is being used to invent spoken claims.
Delivery Pace and pauses marked separately from the script A pause instruction replaces or rewrites a sentence.
Review Transcript, captions and visual reference side by side Captions or lip movement cannot be checked against the source.

For a translated version, the language owner approves the text before mouth movement is judged. A close mouth match cannot establish that a proper name, warning or number was translated correctly. Keep the approved script unchanged while revising only delivery or visual directions.

Keep the identity and speech sources separate

Pexo's talking-avatar workflow describes providing a portrait or character plus a script or message and voice direction, followed by review of delivery, lip sync and appearance. It also calls for permission when using a real person's likeness or voice. The 80-word script and character reference requested in this page's brief are reader-supplied; neither is present here. Record the approved reference and exact speech separately, then compare the rendered face, wardrobe and spoken words with those sources. The public tool description is not a demonstration that this brief produced a consistent presenter.

Review a translated presenter in two passes

First compare the translated script with the approved source: names, numbers, offer terms and warnings must carry the same meaning. A fluent delivery does not establish that those facts survived translation. Second, play the video while checking mouth movement, pronunciation and the blue-jacket host's face and wardrobe against the authorized character reference. If speech is accurate but the face drifts, revise the visual reference; if the face holds but a term is wrong, correct the text or pronunciation cue. Do not use lip sync as a substitute for language approval. This page supplies neither a translation nor a generated presenter.

AI Avatar Video Prompts FAQ

What inputs does a presenter brief need?

Bring an authorized character or portrait reference, an approved script and voice direction.

How should I review translated speech?

Use approved translated wording and ask a fluent reviewer to check pronunciation, names and lip sync.