Hi! I use OpusClip to turn long two-person conversations into Shorts/Reels. I really like the automatic clipping and captions, but the manual editing workflow has become quite frustrating.

My typical workflow is:

long video → choose a specific moment → create a Short/Reel → fix layout/reframing → edit captions → censor words → add an outro → sometimes adjust the clip boundaries later.

There are several issues that make this workflow much harder than it needs to be:

If I have already fixed the layout/reframing, edited captions, censored words and added an outro, and then need to adjust the clip or create another version, previous changes may be lost and I have to repeat the work.

Ideally, manual edits should become the source of truth. If I change the clip boundaries, OpusClip should change the boundaries while preserving my layout, captions, censoring, outro and other manual edits whenever possible.

An option like “Create new version with current edits” would also help.

This is probably my biggest workflow request.

I often already know exactly which part of the long video I want:

03:39 → 05:05

I don't want AI to search for a moment inside that range or decide where the thought starts and ends. I want to enter:

Start: 03:39

End: 05:05

Create clip

and get exactly that section.

Currently, using search/Find Moments may find clips within the requested period rather than simply creating the exact clip I asked for.

The alternative workflow of creating a clip and then using + / Extend a clip / Add section is unnecessarily complicated for such a basic operation.

Ideally there should be a simple manual clipping mode where I can enter timestamps or select IN/OUT points directly on the original video's timeline/transcript.

I sometimes upload a long video because I need its transcript and want to create Shorts from it later.

In one case OpusClip created an almost full-length vertical 9:16 clip from my ~48-minute source video. Because it looked like the processed version of my source, I accidentally published that version to YouTube instead of my original horizontal long-form video.

I think the UI should make a much clearer distinction between:

Original source video

and

AI-generated clips derived from the source.

A full-length source should not silently behave like another generated Short/Reel. It would also be helpful to have an explicit workflow such as “Transcribe only — don't generate clips or change aspect ratio.”

With two-person videos, automatic reframing sometimes moves speakers partly out of frame or crops them incorrectly.

Split layout can also produce inconsistent framing, and manually selecting/tracking two people doesn't work reliably.

Since both speakers in my videos are usually in static video-call windows, I would prefer to simply define two fixed crops:

Speaker A → fixed upper frame

Speaker B → fixed lower frame

and have OpusClip preserve those positions for the whole clip without AI tracking.

Overall

The AI clipping itself is useful. The frustrating part is what happens after the first automatic result.

I would love OpusClip to follow a simple principle:

Preserve everything I changed manually and regenerate only what I explicitly ask to change.

If I change the time range, don't reset my layout.

If I fix the framing, don't reframe it again.

If I edit captions, don't overwrite them.

If I add an outro, keep it.

And if I already know that I need 03:39–05:05, please just let me create 03:39–05:05 directly.

These changes would make OpusClip dramatically easier to use for people who want AI to assist with editing without giving up control of the final clip.

Would love the ability to tell the AI exactly what clip i want, and have it go and add all the magic, cuts, captions etc. I often know what clips i want, and would much prefer a direct request rather than looking through 50 random clips

Quite often IA doesn't not pick the right moments. Would be way better if he had the ability to pick the parts we want out of the whole video.