---
title: "Studio"
description: "Make vertical 1080×1920 videos from your videos and pictures, with hooks, captions, slideshows, before/after, top lists, quote cards and your own music."
canonical: https://postdelegate.com/docs/studio
---

> ## Documentation index
> Fetch the complete PostDelegate documentation index at: https://postdelegate.com/llms.txt
> Use it to find every page before exploring further. Every docs page has a Markdown version at its URL plus `.md`.

# Studio

Make vertical 1080×1920 videos from your videos and pictures, with hooks, captions, slideshows, before/after, top lists, quote cards and your own music.

Studio makes short vertical videos from the videos and pictures you already have. Pick them from your [media library](https://postdelegate.com/docs/media.md), trim, choose what stays in the vertical frame, add text and, if you like, your own music. PostDelegate renders the video on its own servers and saves it to your media library, so it is ready for the composer. Open it from **Studio** in the main menu.

## Templates

| Template | What you get |
| --- | --- |
| Hook over clip | Your clip filling the vertical frame, with a hook line that pops in at the top, middle or bottom. Show it for the whole video or the first 3 or 5 seconds. |
| Talking head + captions | Your clip with burned-in captions that light up word by word. Fill the frame, or keep the whole picture over a blurred copy of itself. |
| Image slideshow | 2 to 10 pictures, each on screen for 1.5 to 5 seconds with its own line of text. Each picture slowly pans, with a fade, slide, wipe or circle transition between them (or a straight cut). |
| Before/after | Two pictures or videos with a label each, one above the other or the second wiping over the first half way through. 2 to 30 seconds. |
| Top-N list | A title over 3 to 7 pictures or clips shown one after the other, each numbered with its label. Count down to #1 or up from #1. A clip plays the part you pick, a picture stays 1 to 10 seconds. |
| Quote card | A quote that appears word by word, with its author, over a colour, a picture or a video. 5 to 15 seconds. |
| AI UGC | A short spoken script over your own footage, with a voiceover and captions that light up word by word. Your AI writes and voices it, or you type it. See [AI UGC](https://postdelegate.com/docs/studio#ai-ugc). |

Every video comes out as an MP4 of 1080×1920 at 30 fps, with H.264 video and AAC sound, which every network accepts.

## Editing a clip

- **Trim**: drag the two handles on the strip of frames. A video can be up to 90 seconds long after speed changes.
- **Crop**: drag the vertical frame over the picture to choose what stays in view.
- **Speed**: 0.5×, 1×, 1.5× or 2×. At 2× you can keep up to 180 seconds of the original.
- **Sound**: keep the sound of the clip or mute it. A muted video still has a silent sound track, because some networks refuse videos without one.
- **Text**: four styles (Bold, Impact, Clean, Classic). The preview uses the same fonts as the render, so what you see is what you get.

Your changes are saved as you go.

Pictures can be PNG, JPG or WebP, up to 20 MB and 25 megapixels (6000 pixels a side). Other picture types in your media library (GIF, AVIF, BMP, TIFF) are left out when you pick them.

## Music

Every template except **Talking head + captions** can have music. Click **Add your music** and upload an MP3, M4A or WAV file of up to 30 MB. Use music you have the rights to: PostDelegate bundles no tracks.

- The music loops if it is shorter than the video and fades out over the last 2 seconds. You can choose where in the track it starts.
- Over a clip with sound, the music sits under it at the level you pick and is lowered further whenever the clip speaks (ducking).
- Every video is brought to the same loudness the networks use (-14 LUFS), with or without music.

Music files are kept for Studio only and never appear in your media library or the composer. A file no project uses any more is removed after 7 days, and it is removed with its project when you delete the project.

## Captions

There are two ways to make captions:

- **With your AI**: if your workspace has its own OpenAI key in **Settings → AI**, Studio sends the sound of the clip to OpenAI's Whisper and gets each word with its timing. Usage is billed to you by OpenAI. See [AI in PostDelegate](https://postdelegate.com/docs/ai.md).
- **With a typed script**: no key needed. Type or paste what is said and Studio spreads the words over the clip, giving longer words more time.

Either way you can fix any word before you render, and choose the caption style, position, words per line and highlight colour.

## AI UGC

AI UGC turns a topic or product into a 15 to 45 second talking-style video over footage you already have.

1. **Footage**: pick up to 6 videos from your media library. Each line of the script gets the next clip, cut to the length of that line. When there are more lines than clips, the clips repeat and pick up where they stopped.
2. **Script**: describe the topic or product, pick a tone (friendly, excited, calm, expert or funny) and a length, then click **Write the script with my AI**. You get a hook, a few lines, a call to action and a post text. Edit any of it before you render. You can also type the whole script yourself.
3. **Voice**: choose a voice and click **Hear the voice** to listen first. The render uses the voice you heard, so you only pay for it once. Pick **No voice, captions only** to render the script as captions over the footage with no voice at all.
4. **AI stills** (optional, with the AI voice): up to 3 images made from your script, shown between your clips with a slow pan.
5. **Music** (optional): add your own music like in the other templates. Under the AI voice it is lowered and ducked whenever the voice speaks.
6. **Render**: Studio voices the script if you have not heard it yet, times every spoken word, makes the stills and renders the video with captions that light up word by word. **Open in composer** starts the post with your post text.

> AI UGC runs on **your own OpenAI key** from **Settings → AI**, and OpenAI bills you for what it uses. PostDelegate never pays for AI and never adds a markup. Studio uses these OpenAI models:
>
> - `gpt-6-luna` writes the script
> - `gpt-4o-mini-tts` makes the voice
> - `whisper-1` times the spoken words
> - `gpt-image-2.5-flare` makes the stills (1024×1536)
>
> OpenAI no longer offers video generation through its API, so AI UGC uses your own footage. If OpenAI refuses a request, Studio shows OpenAI's message. If your key is rejected or your OpenAI account has no credit, Studio tells you so and does not retry.

### Without an OpenAI key

You can still make AI UGC videos:

- **Type the script** and pick **No voice, captions only**. No key is needed.
- **Use Claude or ChatGPT over MCP**: connect your assistant (see [agent files](https://postdelegate.com/docs/mcp/agent-files.md)), ask it to write the script for a Studio project and render it. Your assistant does the writing, and Studio renders the typed script with captions and no voice.

Writing a script and hearing a voice are limited to 30 times an hour, and they need an active plan, like rendering.

## Rendering

Click **Render video**. The video waits for its turn, renders and is saved to your media library, with a progress bar while it works. You can cancel at any time. When it is ready, click **Open in composer** to post or schedule it like any other video.

## Scheduling many videos at once

Under the templates on the Studio page, **Finished videos** lists the videos your projects rendered. Select the ones you want (or **Select all**) and click **Schedule all**:

1. Pick the channels. Fill in any setting a network needs, like a YouTube title.
2. Write the text of the posts.
3. Choose scheduled posts or drafts and confirm.

Each video becomes one post on every channel you picked, in the next free [time slots](https://postdelegate.com/docs/channels.md) of each channel. Every post goes through the same checks as one made in the composer, counts toward your plan's monthly posts, and follows your [approval](https://postdelegate.com/docs/approvals.md) rules. Up to 50 videos and 20 channels at a time. Sending the same selection twice on the same day adds nothing twice.

> Studio is included in every paid plan with fair use:
>
> - one video renders at a time per workspace, and up to 3 more can wait behind it
> - up to 30 videos a day per workspace
> - up to 90 seconds per video
> - the free trial includes 10 videos in total
>
> A render that fails on our side is never counted.

## From Claude, ChatGPT or the API

Your assistant can make Studio videos for you through the [MCP server](https://postdelegate.com/docs/mcp.md), and your own software through the [Studio API](https://postdelegate.com/docs/api/studio.md). Both use the same templates, the same fair use limits and only the videos in your media library.

1. The assistant lists the templates with `studioTemplatesTool`, which describes the inputs of each one.
2. It finds your video with `mediaListTool`, or adds one from a public link with `uploadFromUrlTool`. Links are never used directly in a video.
3. It renders with `studioRenderTool` and waits for the video with `studioRenderStatusTool`. The project also shows up in Studio when you turn on **Show projects made by assistants and the API** under the list, so you can edit it and render it again.
4. The finished video is in your media library. To post it, the assistant schedules it like any other post, which is a draft unless you say otherwise, and your workspace's [approvals](https://postdelegate.com/docs/approvals.md) still apply.

Every template works this way. For captions the assistant writes what is said as a script, and Studio times the words over the clip. For AI UGC it writes the script too, and the video renders without an AI voice unless you ask for one, so no AI key is needed; an AI voice uses your workspace's own OpenAI key, and AI stills come only with the voice. Music must already be uploaded in Studio and not used by another project. Rendering never posts anything on its own, and an assistant can stop a render with `studioCancelTool`.

## What Studio reads

Studio only reads videos and pictures from your own media library and the music you upload. Videos up to 500 MB, 10 minutes long and 4K are supported. Records of renders are kept for 90 days; the videos stay in your media library until you delete them. The fonts are open source (SIL Open Font License) and bundled with PostDelegate.