Talking Portrait Engine

Direct a conversation.

Add talking portraits manually, or turn one prompt into an AI-written video with generated scenes, voiceover, captions, and motion.

Workspace

Toolbox

Talking Portrait

Editor tools

Cast

Your characters

Scene

Background

Video format The canvas, character bounds, and downloaded MP4 use this exact format. Background image (fills the whole video) Use a landscape image for best results. It is included in video exports. External sticker images (emoji, GIF, or image) Each sticker stays visible in Visual Explainer mode, including its blank top area. Stickers remain behind characters and timed captions; GIFs keep animating in supported browsers.

Assets

Built-in PNG library

Category Search

Choose an asset to add it behind the characters, or use its crop icon to add it directly in crop mode. On the stage you can move, resize, rotate, crop, or remove it.

Text overlays

Add styled text to the video

Text

Selected text formattingSelect exact characters or words in the text field, then change a setting below.

Color Font size Font Letter spacing

Bold Italic Underline Strike

Whole text box

Align Line height Background Background color Background opacity

Add the text box first. Select it on the stage, highlight any text in the field, and apply character formatting; box controls affect the complete overlay.

Prompt to timeline

AI Video

Generate styled scenes, cast-aware voiceover, captions, and music automatically.

Separate module

Describe the complete video Include the subject, setting, tone, people, and ending you want. AI will split it into editable scenes.

Production style Caption style Language Video duration Voice engine Voice agent Background music Visuals per scene

Generated scenesDialogue remains editable after generation.

Generated assets are added at the current playhead as full-frame visuals, cast-aware voiceover, optional music, timed captions, and style-matched motion. Existing timeline clips are preserved.

Conversation

Write or perform dialogue

Reel direction

Choose a production style

Cast-powered scenes

Cinematic motion Caption style Responsive placement keeps captions inside the safe area for the selected video format. Render mode captions in preview and export

Each @Character line becomes a cinematic scene. The loaded portrait is reused with camera motion; no scene image is generated.

Script scenes0 scenes

Add mentioned dialogue lines below to build scenes.

Script — one mention per line Character mentions are optional. Matching @Name lines use that portrait; unknown names and plain dialogue are assigned to your available characters automatically. Drag a loaded portrait anywhere on the stage.

Lip Sync Engine Dynamic Hindi (recommended) Basic fallback

Voice and language are selected automatically from the script—no API key is required. Dynamic Hindi aligns mouth shapes to generated or uploaded dialogue audio; if automatic voice is unavailable, preview continues with browser speech and text-driven lip sync.

Checking your ITRSP Lipsync access…

Each script line is placed automatically on its character track with synchronized voice and captions. Uploaded dialogue audio remains available for manual replacement.

Sequence editor

Timeline

Arrange every scene element with frame-accurate control.

＋ Audio ＋ Music ＋ SFX ＋ Video ＋ Image

Select any clip to enter an exact start, duration, or end. Dynamic character pose clips control position and face style for their time range. Drag clips to move them and drag either edge to trim. Drag dialogue vertically between character lanes to change its speaker. Shift-click selects multiple clips. Snaps to 30 fps.

Add a character image to begin. The first load may download the face-landmark model.

Direct a conversation.

Choose a production style

Timeline

Timeline object

Log in to ITRSP