Auto B-Roll: Add Stock Footage From Your Captions

Quick answer: Add captions first - Auto B-Roll works from them - then let Zella pick the most visual keyword per line using on-device text analysis and drop cutaways across your video, spaced out rather than every N seconds. It needs no key and no account on any platform. With no key, Zella draws a designed keyword card for each line entirely on your own device. Add a free Pexels API key (pexels.com/api, stored in the macOS Keychain) only if you would rather have filmed stock footage. Only ever a keyword leaves your device; never your video. Auto B-Roll is free on every tier.

On this page: what it is · add captions first · run it · what it places · swap, trim, remove · optional Pexels key · privacy · on iPhone · FAQ

The steps below are the Mac flow. For the touch version, jump to On iPhone.

What is Auto B-Roll?

B-roll is the extra footage you cut to while someone keeps talking - a shot of a city, a laptop, a coffee cup - instead of staring at the same face the whole time. It makes a video feel edited and keeps people watching.

Auto B-Roll finds and places that footage for you. It reads your captions (the written transcript of what you said), figures out the most visual word in each line, finds a matching free stock clip online, and lays it onto your timeline as a short cutaway. What used to be a manual chore - pause, search a stock site, download, drag, trim, repeat - becomes one click.

It lives in two places in the editor:

The AI panel has an Add B-Roll button.
The AI Assistant can do it when you ask, for example: “add b-roll based on my captions.”

Step 1: Add captions first

Auto B-Roll works from your captions, so you must have captions before you run it. If you skip this, there’s nothing for it to read.

In the editor, open the Captions tab and run Enable Captions / Transcribe audio (see the captions guide), or
Ask the AI Assistant to add captions first.

Once your video has a word-level transcript, you’re ready.

Step 2: Run Auto B-Roll

Pick whichever way you prefer:

Button: open the Auto panel and click Add B-Roll.
Assistant: type something like “add b-roll based on my captions.”

Zella then, on its own:

Reads each caption line.
Picks the most visual keyword in that line using on-device NaturalLanguage - Apple’s built-in, on-your-device text analysis. Nothing but the derived keyword ever leaves your device.
Draws a designed keyword card for that word - a dark gradient card with the word set on it - and places it on the timeline as a cutaway.

The cutaways are spaced out across your whole video, not dropped blindly every few seconds, so the b-roll feels intentional instead of mechanical.

No key, no account, no network. This is the default path and it runs entirely on your device. The cards are labelled as keyword cards in the status line, never presented as stock footage, so you always know which you are looking at.

Optional: add a free key for real stock footage

Keyword cards carry the point of a line, but they are typography, not footage. If you want filmed shots instead, Auto B-Roll can fetch them from Pexels, a library of free stock video.

Go to pexels.com/api and sign up (it’s free).
Copy your API key.
Back in Zella, run Auto B-Roll. The first time, Zella prompts you once to paste the key.
Paste it in. Zella stores it only in the macOS Keychain - the same secure vault your Mac uses for passwords - so you never have to enter it again.

Pexels clips are free to use, including commercially, and usually need no attribution. There’s no extra cost from Zella and no subscription for this feature.

This is the one step that needs the internet, because Zella is fetching clips from Pexels. Only the derived keyword is sent. Your recording and your footage never leave your device.

What Auto B-Roll places

Each cutaway Zella adds is:

Full-frame - it fills the screen, covering your main footage for the length of the clip (a true cutaway, not a small picture-in-picture).
Silent - the stock clip’s own audio is off, so your narration keeps playing underneath.
About 4 seconds long - short enough to keep the pace moving.
Spaced across the video - distributed over the whole timeline, tied to the caption lines with the most visual keywords.

Everything lands as a normal overlay on the timeline, so you stay in full control of it.

How to swap, trim, or remove a clip

A placed b-roll clip behaves like any other overlay - nothing is locked in:

Swap it: don’t like the clip Zella chose? Remove it and drop a different one in, or add your own footage in the same spot.
Trim it: drag the edges on the timeline to make the cutaway shorter or longer.
Move it: slide it earlier or later so it lands on the exact word you want.
Remove it: delete it like any overlay, and your main footage plays through underneath as normal.

Run Auto B-Roll for a fast first pass, then tidy up the handful of clips you want to change.

Is Auto B-Roll private?

Auto B-Roll is opt-in and clearly network-based, and it’s careful about what it sends:

Only the keyword leaves your device. Zella derives a single visual word from each caption line, on-device, and sends that word to Pexels to search. Your video is never uploaded.
Keyword extraction is on-device. The NaturalLanguage step that picks the word runs entirely on your device.
Recording and your own footage stay 100% on-device. Only the b-roll search and download use the internet - because fetching free stock clips has to.
Your Pexels key is stored only in the macOS Keychain, never in a plain file.

If you’d rather not fetch anything from the internet, simply don’t run Auto B-Roll - the rest of Zella keeps working entirely on-device.

On iPhone

The phone works the same way the Mac does: b-roll with no key, no account and no network at all.

The default: keyless cutaway cards

Add captions first - Auto B-Roll works from them, exactly as on the Mac.
Tap Add cutaways.
Zella picks the most visual keyword per caption line on-device, generates a cutaway card for each - a designed gradient card with the word on it - and drops each one onto the timeline near the line that mentions it.

These are generated motion-graphic cards, not licensed stock footage, which is exactly why they need no key and never touch the network. The whole run is 100% on-device. The Mac does the same thing when no key is set, so neither platform asks you to sign up for anything before Auto B-Roll will run.

Optional: real stock footage

Want actual stock clips instead? Open Use my own stock footage (Pexels / Pixabay), paste a free key from either provider, and tap Add stock b-roll. Zella stores the key in the device Keychain, and only the keyword leaves the phone to search - your video never does. This is the same path the Mac uses, with Pixabay as an extra provider option.

How the two platforms compare

	Mac	iPhone
Default source	Generated keyword cards	Generated cutaway cards
Key needed to start	None	None
Network needed to start	No	No
Optional providers	Pexels	Pexels or Pixabay
What leaves the device	One keyword per line, and only on the optional path	One keyword per line, and only on the optional path
Price	Free	Free

Full detail: on-device AI on iPhone.

Auto B-Roll FAQ

Do I need captions before I can add b-roll? Yes. Auto B-Roll reads your captions to pick keywords, so run Enable Captions (or ask the Assistant to add captions) first.

Where do the clips come from, and are they free? From Pexels, a free stock-video library. The clips are free to use, including commercially, and usually need no attribution.

Do I have to pay for the Pexels key? No. A Pexels API key is free - sign up at pexels.com/api, copy the key, and paste it into Zella once.

Does my video get uploaded anywhere? No. Only the keyword derived from each caption line is sent to Pexels to search. Your video never leaves your device.

How long are the b-roll clips, and do they have sound? They’re about 4 seconds each, full-frame, and silent, so your narration keeps playing underneath.

Can I change the clips Zella picked? Yes. Each clip is a normal overlay - swap it, trim it, move it, or delete it like any other.

Why is b-roll the one internet feature? Because fetching stock footage requires the internet. Recording and editing your own footage stay 100% on-device; Auto B-Roll is opt-in. On iPhone, the default cutaway cards need no network at all.

Do I need a key for Auto B-Roll on iPhone? No. Add cutaways generates gradient cutaway cards on-device with no key and no account. A free Pexels or Pixabay key is only for the optional real-stock-footage path.

Is Auto B-Roll a Pro feature? No - it is free on both platforms, like all of Zella’s AI including the Assistant.

Why does the phone place cards instead of stock clips? Because a generated card needs no licence, no key and no network - so it works offline and out of the box. Add a free Pexels or Pixabay key if you’d rather have real footage.

Pro tips & gotchas

Captions first, always. No captions means no keywords to search, so Auto B-Roll has nothing to work with.
A key is optional. Without one you get keyword cards. Enter a Pexels key once for filmed footage and it is saved in the Keychain; you won’t be asked again.
Treat it as a first pass. Let Auto B-Roll place clips across the video, then swap or trim the few you want to change.
You need to be online for this step only - it’s fetching clips from Pexels. On iPhone the default cutaway cards need no network at all.
On the phone, tap Add cutaways to start. No key, no account, nothing to sign up for - add the optional Pexels or Pixabay key only if you want real stock footage.