Opus Clip

Opus Clip is the tool we use when we have a 45-minute podcast or interview that we want to extract 8-12 shorts from. The clip-selection AI watches the long-form video, finds moments that have hook structure (setup-payoff inside 30-90 seconds), and exports them with auto-captions and proper vertical framing.

It is not as fully automated as the marketing pitch suggests, but the time savings on the clip-identification step are real.

What it does

Upload a long-form video or paste a YouTube URL. Opus Clip transcribes it, identifies candidate clips ranked by a "virality score," and exports each candidate as a captioned 9:16 vertical video.

Output: 5-15 short clips, each 15-90 seconds, ready to post to TikTok, Shorts, or Reels.

Where it wins

Clip selection on dialogue. For podcasts, interviews, and talking-head explainers, the clip selection is 60-70% "actually a good short" out of the box. That's a strong hit rate compared to manually scrubbing the timeline for hook moments.

Auto-captioning. Captions are auto-generated and styled in the clip output. The styling is more generic than Submagic's, but for "just ship it" speed, it's fine.

Multi-platform sizing. Outputs in 9:16 vertical by default. Can also export 1:1 square and 16:9 landscape from the same source. Saves the "resize for every platform" step.

Virality score. The score isn't magic but it's directionally useful for triage. A score above 80 is usually worth posting. Below 60, skip it. The 60-80 band is where the operator judgment matters.

Where it underdelivers

Clip quality on non-dialogue content. If the long-form video is graphics-heavy or has long stretches without spoken hooks (think product reviews, gameplay, screen recordings), Opus struggles. The clips it picks tend to be the dialogue moments, missing the visual hooks the long-form was actually built around.

Caption styling. The default captions look fine but generic. If you want the "cinematic short" styling that's converting on TikTok in 2026, run the Opus output through Submagic for the caption layer.

False positives. Roughly 40% of the suggested clips aren't actually short-worthy. You still need a human pass to filter, just a much shorter pass than scrubbing the original video.

Pricing reality check

The $9/mo Starter tier gives you 90 minutes of video processing per month, which is one or two long-form videos' worth of source material. Enough to test the workflow.

The $19/mo Pro tier (300 min/mo) is the realistic working tier for an operator running multiple long-form channels with short-form repurposing.

Free tier exists but watermarks output. Skip it for serious use.

Stack fit

We pair Opus Clip with:

• Descript when we want to do the long-form cut first, then auto-clip the cut (cleaner than auto-clipping the raw)
• Submagic for the caption styling layer on selected clips
• Manual export to YouTube Shorts, TikTok, and Reels via native platform schedulers

The typical workflow: ship the long-form cut, paste the YouTube URL into Opus, pick 5-8 of the suggested clips, run them through Submagic for styling, schedule across platforms.

Should you use it

Yes if:

• You publish long-form content (podcasts, interviews, explainers) and want short-form as a derivative output
• You've been manually scrubbing your timeline for clip-worthy moments and want to automate that
• Your long-form is dialogue-driven

No if:

• Your long-form is graphics-heavy or non-dialogue
• You're already shipping shorts as a primary format (Opus is for repurposing, not native short creation)
• You need fully custom clip selection beyond what an AI can infer

Try it

Try Opus Clip free→

Disclosure: affiliate link. Commission on paid upgrades. We use Opus Clip on long-form-to-shorts repurposing workflows.