video-generator

Opus Clip

Name: Opus Clip
Availability: InStock
Author: AI Demos

Hands-off vertical clip generation with captions and emoji accents, if you can live with preset styling and paywalled exports.

Visit Opus Clip

9:16 auto-cropSemantic emojisWatermarked exportPaywalled fonts/SRT

TL;DR — our verdictUpdated July 2026 · 7 test artifacts

Strong automation for repurposing, but not a full caption-design solution

Where it wins

you want to auto-crop a source video into a face-centered 9:16 short with minimal manual work
you want semantic emoji overlays and ready-made caption styling
you want clip scoring and scene analysis to find highlights quickly

Main limitation

you need custom font uploads or hex color control

Pricing (verified plans)

Starter $16 USD / $9 USD/moPro $29 USD / $9.5 USD/moBusiness Let's talk

Strongest test artifacts

Auto Short-Form Repurposing with Face Tracking →Scene Analysis and Clip Scoring →Kinetic Caption Styling and Timing →

Our take

Opus Clip is a strong fit when you want a source video turned into a clean vertical short with minimal manual work. It did the face-centered crop well, added semantic emojis, and stayed stable through pauses, but fast speech exposed caption lag and the freemium workflow blocks the custom fonts, hex colors, and raw SRT exports that brand-heavy teams usually need.

Tutorial walkthrough of the Opus Clip workflow, clip review, and export controls.

In-Depth Review

Our detailed analysis of Opus Clip — features, performance, and real-world testing.

AI Demos Team

Expert Reviewer

Verified Review

Feature-by-Feature Breakdown

Auto Short-Form Repurposing with Face Tracking

Strong

▾

Test Summary

Feature tested: Auto Short-Form Repurposing with Face Tracking

Result: Passed — Strong

Opus Clip repurposes a landscape source into a 9:16 short while keeping the speaker centered. In the markdown-pages test and the narrative clip test, the output stayed locked on the face through minor movement and preserved a clean 1080p profile.

INPUT

Source video: "Ai Demos now supports markdown pages - SEQ.mp4" — a 16:9 talking-head clip with technical markdown terminology, including the literal ".md" extension.

↓→

image

Output artifact for "Auto Short-Form Repurposing with Face Tracking" test: The tool auto-reframed the 16:9 source into a vertical 9:16 clip, kept the speaker centered, and displayed a 90/100 clip score with Hook, Flow, Value, and Trend grades. The report says the export stayed clean at 1080p, but the transcription flattened technical syntax by turning ".md" into lowercase markdown and omitting the period., output-1.png

The tool auto-reframed the 16:9 source into a vertical 9:16 clip, kept the speaker centered, and displayed a 90/100 clip score with Hook, Flow, Value, and Trend grades. The report says the export stayed clean at 1080p, but the transcription flattened technical syntax by turning ".md" into lowercase markdown and omitting the period.

INPUT

Source video: "Workflow vs AI Agent - SEQ Copy 01.mp4" — a conceptual RAG vs. AgentiGRAG segment with narrative pauses and a longer speaking arc.

↓→

image

Output artifact for "Auto Short-Form Repurposing with Face Tracking" test: The clip review screen showed a vertical segment view with subtitles and an 86/100 score. The report says the tool parsed the narrative structure smoothly, kept the reframed view stable, and avoided timeline fractures while handling the longer explanation., output-3.png

The clip review screen showed a vertical segment view with subtitles and an 86/100 score. The report says the tool parsed the narrative structure smoothly, kept the reframed view stable, and avoided timeline fractures while handling the longer explanation.

Bottom Line

Reliable for hands-off repurposing into portrait shorts; face-centering and clip packaging were strong, but it is not a precision transcriber for technical syntax.

Scene Analysis and Clip Scoring

Strong

▾

Test Summary

Feature tested: Scene Analysis and Clip Scoring

Result: Passed — Strong

The editor exposes a scene-analysis view with transcript text and clip scores so users can triage highlights quickly. In the observed outputs, one clip was scored 90/100 with Hook, Flow, Value, and Trend grades, and the narrative clip was also shown with a score in the review screen.

INPUT

Source video: "Ai Demos now supports markdown pages - SEQ.mp4" — used to test automatic clip selection and score breakdown.

↓→

image

Output artifact for "Scene Analysis and Clip Scoring" test: The scene-analysis page showed a 90/100 score and broke the clip into Hook, Flow, Value, and Trend grades. The transcript panel and selected clip view make it clear that Opus Clip is doing automated highlight triage, not just passive captioning., output-1.png

The scene-analysis page showed a 90/100 score and broke the clip into Hook, Flow, Value, and Trend grades. The transcript panel and selected clip view make it clear that Opus Clip is doing automated highlight triage, not just passive captioning.

INPUT

Source video: "Workflow vs AI Agent - SEQ Copy 01.mp4" — used to test narrative clip scoring and scene analysis on a more conceptual segment.

↓→

image

Output artifact for "Scene Analysis and Clip Scoring" test: The rendered review screen visibly shows an 86/100 score for the RAG vs. AgentiGRAG clip. The report text separately mentions a 91/100 Viral Score, so the exact number differs by view, but the tool is clearly surfacing clip-level scoring and transcript-based scene analysis., output-3.png

The rendered review screen visibly shows an 86/100 score for the RAG vs. AgentiGRAG clip. The report text separately mentions a 91/100 Viral Score, so the exact number differs by view, but the tool is clearly surfacing clip-level scoring and transcript-based scene analysis.

Bottom Line

Useful for clip triage, but the report text and the rendered screen disagree on the narrative clip's exact score (91/100 vs 86/100).

Kinetic Caption Styling and Timing

Mixed

▾

Test Summary

Feature tested: Kinetic Caption Styling and Timing

Result: Partial — Mixed

The caption engine burns in styled subtitles, can place semantic emojis above matched phrases, and keeps subtitle cards visible through pauses. On the chatbot clip and the 1.15x playback stress test, it rendered the cue styling correctly but showed timing drift under faster speech.

INPUT

Source video: "AI demos chatbot short.mp4" — a fast-paced monologue used to test semantic emoji placement and caption timing under acceleration.

↓→

image

Output artifact for "Kinetic Caption Styling and Timing" test: The clip review screen showed a 90/100 score and the report says the tool successfully matched semantic emojis to spoken phrases such as "discover." Under a 1.15x stress test, the kinetic text tracking lagged slightly and produced minor frame drops, though the output remained readable., output-2.png

The clip review screen showed a 90/100 score and the report says the tool successfully matched semantic emojis to spoken phrases such as "discover." Under a 1.15x stress test, the kinetic text tracking lagged slightly and produced minor frame drops, though the output remained readable.

INPUT

Source video: "Workflow vs AI Agent - SEQ Copy 01.mp4" — used to check whether captions stay stable through pauses and conversational breaks.

↓→

image

Output artifact for "Kinetic Caption Styling and Timing" test: The report says the silence-detection logic isolated breathing spaces and conceptual transition frames without fracturing the timeline, and the text stayed on screen through empty audio windows until the next phoneme began. That makes the timing engine patient on pauses, even if it is less precise on very fast speech., output-3.png

The report says the silence-detection logic isolated breathing spaces and conceptual transition frames without fracturing the timeline, and the text stayed on screen through empty audio windows until the next phoneme began. That makes the timing engine patient on pauses, even if it is less precise on very fast speech.

Bottom Line

Good for template-driven kinetic captions, but rapid speech still exposes visible timing drift.

Export and Brand Customization Controls

Poor

▾

Test Summary

Feature tested: Export and Brand Customization Controls

Result: Failed — Poor

The export path can deliver MP4s, but in the freemium workflow the output was watermarked and the interface blocked custom font uploads, hex color control, and raw SRT downloads. The paywall modal also advertised higher-tier XML export and multiple aspect ratios.

INPUT

Attempt to change brand parameters and export cross-compatible subtitle files on the observed freemium workflow.

↓→

image

The upgrade modal shows Starter, Pro, and Business tiers. The report says free-tier exports are watermarked and that custom font uploads, hex colors, and raw SRT downloads are blocked behind paid tiers.

Bottom Line

Free-tier output is usable only as a review asset; brand customization and subtitle portability are paywalled.

Observed plan tiers

The modal presents Starter, Pro, and Business, with paid tiers unlocking brand and export controls.

Starter

$16 USD / $9 USD/mo

150 credits; AI clipping with Virality Score; AI animated subtitles in 20+ languages; auto post to YouTube Shorts, TikTok, IG Reels, or download; powerful editor; 1 brand template; filter & silence removal; remove watermark.

Pro

$29 USD / $9.5 USD/mo

3,600 credits per year, available instantly; team workspace with 2 seats (up to 4); 2 brand templates (up to 4); 6 social media connections (up to 12); AI B-Roll; input from 10+ sources; export to Adobe Premiere Pro & DaVinci Resolve; multiple aspect ratios (9:16, 1:1, 16:9); social media scheduler; Intercom chat support; custom fonts; speech enhancement; download subtitles, transcripts & text tracks.

Business

Let's talk

Everything in Pro plus priority project processing; customized credits and team seats; tailored business assets; dedicated storage; API & custom integrations; Master Service Agreement; priority support with a dedicated Slack channel; enterprise-level security; download subtitles, transcripts & text tracks.

Pricing and inclusions were observed in the upgrade modal as displayed.

✓ Use This If

●you want to auto-crop a source video into a face-centered 9:16 short with minimal manual work

●you want semantic emoji overlays and ready-made caption styling

●you want clip scoring and scene analysis to find highlights quickly

✕ Skip This If

●you need custom font uploads or hex color control

●you need raw SRT downloads or other sidecar subtitle exports on the observed tier

●you need a watermark-free free-tier export

●you need pixel-tight handling for technical punctuation or very fast speech

video-generatorshort-form-video-assistantvideo

Yes. In the markdown-pages test, it accepted a 16:9 source and produced a 9:16 vertical clip with the speaker face-centered. The output stayed clean at 1080p.

Not well. The report says it flattened technical syntax into lowercase text and omitted the period in ".md".

Yes. In the fast chatbot clip, it placed semantic emoji-style cues above matched phrases, including pairing an investigative cue with the word idea of "discover".

Yes. The report says the freemium export is watermarked, and the watermark is permanent on the final video track.

No. The report and the paywall modal both say custom font uploads, hex color control, and raw SRT downloads are blocked behind paid tiers.

The upgrade modal shows Starter, Pro, and Business. Starter is shown at $16 USD / $9 USD/mo, Pro at $29 USD / $9.5 USD/mo, and Business says "Let's talk."