How It Works

1
Tell the Agent your idea
Type a concept, paste a script, upload a PDF, or describe the ad you want. No prompt engineering required — the Agent asks clarifying questions if it needs more detail.
2
Watch the plan take shape
CrePal's AI Director breaks your idea into scenes, writes the shot-by-shot script, and selects the best-fit model for each shot — a realistic model for live-action shots, a stylized one for animation, a dedicated lip-sync model for talking heads. You see this plan before generation starts.
3
Review and redirect in real time
As scenes generate, you watch them populate live. Don't like a shot? Tell the Agent — "make the lighting warmer," "replace this scene," "shorten the intro" — and it revises without you touching a timeline.
4
Export production-ready video
Final output includes synced subtitles, voiceover, and a mixed soundtrack, exported in HD and ready to publish.

Why an Orchestration Layer (and Not Just Another Model)

Image and video generation models are good at narrow things: one model renders photoreal scenes well, another handles motion consistency, another does lip-sync, another scores music. Picking the right one for each shot — and stitching the outputs into a coherent video — is itself a skill.

CrePal's AI Director automates that decision. It currently coordinates models, including Runway, Kling, Veo, Hailuo, PixVerse, Vidu, Recraft, Suno, and OpenAI's image/video models, routing each part of your script to whichever model produces the best result for that specific shot type. You don't choose a model; you describe an outcome.

Built for How You Actually Work

AI Story Mode — for first-time and casual creators

Describe what you want once. The Agent handles scriptwriting, scene planning, model selection, generation, and editing end-to-end. You review the result and request changes in chat.

Example: a small business owner with no editing background describes a 30-second product ad. The Agent drafts the script, generates four scenes, adds a voiceover and subtitles, and delivers an export-ready file — no timeline, no manual cuts.

Director Mode — for creators who want control

Same orchestration engine, but with shot-by-shot editing, model overrides per scene, and direct parameter control (camera motion, pacing, aspect ratio) for users who know what they want and want to fine-tune it.

Example: a freelance motion graphics artist requests "shorter intro, punchier cuts, warmer palette" through chat and gets a revised cut without reopening a timeline editor.

Mini Apps — for repeatable, specific jobs

Purpose-built flows for common formats: turning a PDF or deck into a narrated video, generating a talking-avatar spokesperson video from a script, or remixing an existing ad's style around your own product. Each Mini App skips the general planning step and goes straight to the specific output.

Example: a marketing manager uploads a PDF case study and gets a narrated explainer video in minutes, without writing a script from scratch.

What's Included

Capability	What it does
Multi-model orchestration	Automatically selects from 10+ leading video, image, voice, and music models per shot
AI Director planning	Breaks ideas into scripted, scene-by-scene production plans before generation
Conversational editing	Revise scenes, pacing, tone, and style using natural-language chat instead of a timeline
Real-time preview	Watch scripting, generation, and assembly happen live, with the ability to interrupt and redirect
Subtitles & voiceover	Auto-generated, synced, and editable in the final export
HD export	Production-ready files with audio mixing included

One Agent. Every Step of Video Production.

Have a workflow you don't see covered here? Tell us what you're building — Mini Apps are added based on real creator requests.

Get Started for Free

How It Works

Tell the Agent your idea

Watch the plan take shape

Review and redirect in real time

Export production-ready video