---
name: make-podcast-video
description: Create one complete podcast video from approved audio or script with speaker-aware visuals and synchronized program audio. Use for full podcast episodes or excerpts that need host, guest, captions, and program structure; do not use for a single talking portrait, generic social montage, music video, or audio-only delivery.
---
# Make Podcast Video

## Catalog

- Display name: Make Podcast Video
- Categories: Video
- Status: Ready
- Route label: Podcast scene and audio-led video assembly
- Tone: navy
- Short description: Create one complete podcast video from approved audio or script with speaker-aware visuals and synchronized program audio.

- Cover image: /skill-covers/make-podcast-video.png

## What this skill does

- Builds a finished podcast video instead of returning loose talking clips.
- Maps speaker turns, titles, captions and supporting visuals into one coherent timeline.

## How to use

Provide approved audio or script, speakers and consent, show identity, platform, runtime and captions.

#### Example

```text
Create a 30-second 16:9 podcast highlight from authorized two-speaker audio; preserve wording and identify speakers.
```

## Workflow

1. Verify audio rights and speaker consent; identify turns, key quote and placement.
2. Choose real footage, authorized avatars or a faceless visual system; create an audio-led plan.
3. Generate only missing visuals: use GPT Image 2 Medium/2K for stills with readable text or consistency needs, Kling 3.0 to animate a precise supplied still, and Seedance 2.5 only when a generated shot requires audio-visual synchronization. Then assemble every scene deterministically to the approved audio.
4. Watch end to end; verify wording, attribution, sync, captions, continuity, loudness and assembly.

## User-facing output

- Media type: One MP4 video
- Default quantity: 1
- Content per video: One complete audio-synchronized podcast video
- Default layout: User-requested delivery format
- Model policy: Deterministic audio-led assembly; GPT Image 2 Medium/2K for stills, Kling 3.0 for precise still animation, Seedance 2.5 for generated audio-visual shots
- Downstream use: Podcast episodes and social highlights