GTA 6 clip with webcam on top and gameplay below: how to set it up
Streamer cam on top and gameplay below is the format that shows the reaction next to the play in GTA 6 clips. Here's what Cut.Pro's AI does with each webcam shape, how much of the game is left in vertical, and the step-by-step to get it ready in a template.

GTA 6 clip with webcam on top and gameplay below: how to set it up
To build a GTA 6 clip with the webcam on top and the game below in Cut.Pro, turn on automatic reframing in your template and check the "Webcam and screen" layout: the AI finds the camera in the corner of the stream and builds the two strips on its own. The shape of the streamer's webcam decides how much of the game fits in the vertical video. Below, the math, the step-by-step and the mistakes that make a clip miss the play.
For the GTA 6 streams that started with Thursday's (19th) launch, this is the format that makes sense: whoever pays for clips, whether a clipping campaign or a championship, pays for the streamer's reaction, and the reaction has to show up next to what caused it. A clip with only the game looks like every other profile's. A clip with only the face doesn't show the joke.
What does the AI do with a GTA 6 stream?
It looks for a still face in a corner of the screen and, if it finds one, crops that corner into the top strip and a slice of the game into the bottom one. If it doesn't find one, it falls back to one of the base framings.
Three behaviors matter for anyone clipping GTA:
- It only splits when the camera is in the corner. If the stream opens with the streamer full screen talking to chat, that stretch comes out as "Full screen," and the clip only splits when the camera goes back to the corner over the game.
- The game slice avoids the camera. If the webcam sits near the middle of the screen, the bottom slice shifts sideways so the face doesn't show up twice. The crosshair and the car can end up off-center in the clip.
- A camera with two people counts too. A duo playing side by side on the same webcam goes into the top strip as a single block.
How much of the game is left in vertical?
Between roughly 42% and 63% of the game screen's width, on a 16:9 stream. The webcam shape decides.
The top strip copies the camera's aspect ratio to fill the video's width with no bars, limited to a minimum of 22% and a maximum of half the height. The game takes the rest. The taller the bottom strip, the more the game is zoomed in and the less of its width fits. I ran the numbers for the most common camera shapes:
| Webcam shape on the stream | Camera strip (video height) | Game width that shows |
|---|---|---|
| Ultrawide (21:9) | about 24% | about 42% |
| Wide (16:9) | about 32% | about 46% |
| Classic (4:3) | about 42% | about 55% |
| Square (1:1) | 50% | about 63% |
In practice: a streamer with a wide camera produces a clip with the game big and close, good for chases and fights, where the action stays in the middle. A streamer with a square camera produces a clip with more scenery on the sides, but the game is smaller, and small in-game text is hard to read on a phone. Either way, the corners of the game screen are left out, and corners are where GTA games tend to put the radar and indicators. If the joke depends on the map or the wanted level, write the context as text in the clip.
This also explains the rule the blog already covered in captions for GTA 6 clips: the streamer's captions go on the line between the strips, and that line moves with the camera. One template per streamer.
Step-by-step in Cut.Pro
The layout lives in the template, not in the clip. You set it once and every clip generated with that template already comes out split.
- Open the template. In the menu, under "Templates," open yours (the card says "My template"). If you don't have one yet, open any clip in the editor and use "Save as template" in the editor menu.
- Turn on reframing. In the "Layout" tab, enable "Auto reframe." The description on screen says what it does: an AI follows who is speaking and reframes the video to the chosen format.
- Check the layouts. Under "Allowed layouts," leave "Webcam and screen" and "Full screen" checked. The AI picks among the checked ones depending on the stretch, and one of the two base layouts ("Full screen" or "Whole frame") is always on. For a GTA stream, I'd uncheck "Webcam and video" and "Split screen": both were made for real people in the bottom video, and GTA cutscenes are full of character faces.
- Choose the framing. Under "Framing," "Fit" respects the template's video box, which is useful if you added a frame, a brand or a fixed call to action. "Free" takes up the whole screen.
- Put the template on the stream. In "Clip live," open the monitored channel and, under "Caption style and look," click "Change style" and choose the template. For a VOD or uploaded gameplay, the same field shows up in "Clip video" before you generate.
- Check the first three. Open the first clips and see whether the face is fully inside the top strip and whether the action landed in the bottom slice. If the streamer's webcam moves between scenes, check stretches from different scenes.
If you're not monitoring the channel yet, the "Add channel" and "Monitor" path is in the launch night plan, and how many credits that uses per hour of stream is in clipping a GTA 6 live without burning credits.
Why not use the simple editor's "Split"?
Because the fixed formats crop the middle of the image. In the simple editor, the "Video layout" section has "Split" and "Webcam," and both build the strips from the center of the frame. On a GTA stream with the camera in the corner, the top strip shows the center of the game, not the face.
The fixed formats are for something else: video where what matters is already in the middle. For your own gameplay recorded on a console, with no camera, the path is different and is in how to clip GTA 6 gameplay recorded on PS5 or Xbox. There, "Whole frame" (the full screen in a strip, with a blurred background) keeps the radar and the screen corner that the vertical slice would cut.
Common mistakes
- One template for three streamers. Each camera produces a strip of a different height, and captions positioned for one land on top of the other's face.
- Leaving every layout checked. The AI is more likely to pick a conversation framing in a cutscene stretch. Check only what works for gameplay.
- A clip that depends on the screen corner. A money glitch with the counter in the corner, a chase that only makes sense with the radar: the vertical slice doesn't show it. Write the number on screen or pick another moment.
- Posting a whole cutscene just because the layout looks good. Rockstar's policy lists isolated cinematic scenes among what can be taken down. The webcam layout helps precisely because it puts the streamer's reaction up front, not the game scene.
If you want to go beyond the layout, there are two neighboring posts: how to get the AI to find the moments that become clips and one GTA 6 clip on every network, which shows what changes from network to network once the clip is ready. For the general side of the topic, outside GTA, see the guide on how to cut gameplay into vertical clips.
Where to start
Today, look at the camera of each streamer you clip and note its shape. A wide camera calls for a template built for action in the center; a square camera calls for attention to small in-game text. Create one template per streamer with "Webcam and screen" and "Full screen" checked, apply it to the stream and check the first three clips before scheduling. With 125 million hours of GTA 6 streams watched in the first 24 hours, according to the Streams Charts recap (details in who to clip right now), there will be no shortage of material.


