How to Get a Frame Every Second from a YouTube Video

Last updated: 7 November 2026 · By the FrameCap team

Set FrameCap's time between frames to 1 second. You get one picture for every second of the part you choose, each with its time, as a PDF for AI chatbots or a ZIP for AI agents. One second suits fast screens, like code being typed.

When one frame a second makes sense

Use 1 second when the screen changes fast: code typed live, quick clicks through settings, or a short demo. For slides and talks, a frame every 5 to 10 seconds is enough.

How to do it

Open the YouTube video and click the FrameCap button on it.
Move the time between frames to 1 second.
Pick the part you need, for example 2:30 to 4:00.
Click to capture.

How many frames that is

One minute gives 60 frames. Ten minutes give 600. That's a lot for a person or an AI to read, so pick only the part you need.

Fewer repeated pictures with Smart frames

At 1 second, many pictures look the same, for example while someone talks over one screen. FrameCap Pro's Smart frames keeps a new picture only when the screen changes, so you get the moments that matter.

What you get

For AI chatbots: a PDF of the frames, the transcript and a ready-made prompt. For AI agents: one ZIP with pages of frames (2, 4 or 6 per page in the free version, up to 24 in FrameCap Pro), the transcript and a README.

FAQ

Can I get every single frame of a video?

No. FrameCap takes a picture every 1 to 30 seconds, which is what you need to follow a video. One picture a second is the most detailed setting.

Is one frame a second too much for ChatGPT?

For a long part, it can be. Use 1 second for short parts and 5 to 10 seconds for long videos.

How long can the video be?

Up to 40 minutes at a time in the free version, and up to 3 hours in FrameCap Pro.