Can AI clip a Nepali-language podcast into shorts?

ClipNepal transcribes Nepali audio directly, scores every candidate moment, and renders vertical clips with the captions burned in. The transcription is prompted for Nepali in Devanagari rather than romanised, and English loanwords are kept in Latin script where the speaker used English.

Why do Nepali captions render as empty boxes or broken letters?

Most caption fonts have no Devanagari coverage, so they draw a row of empty boxes, and most subtitle renderers do no shaping, so conjuncts decompose. Captions have to be rendered by something that shapes Devanagari and in a font that carries the Devanagari block, including the virama that forms every conjunct.

Which fonts can set Devanagari captions?

Khand, Teko and Mukta are display faces with full Devanagari coverage, and Noto Sans Devanagari covers the script in weights up to Black. Khand and Teko draw the Devanagari digits as Western figures, so ५० renders as 50 in either of them.

What does an Opus Clip alternative for Nepali need that an English one does not?

A transcriber prompted for Nepali rather than auto-detect, a caption renderer that shapes Devanagari conjuncts, fonts that carry the script, and clip boundaries that land on Nepali word edges rather than mid-word.

Dual-Speaker Layout

Create stacked split-screen clips for two-person podcasts and interviews.

In wide-angle interviews, a single vertical crop cuts out the other speaker. ClipNepal detects both faces, verifies distinct positions, and stacks them into high-retention 2-up vertical video.

No credit card requiredHarfBuzz Devanagari typography1080p full HD export

Core Capabilities

Engineered specifically for short-form virality

Multi-Face AI

Automatic Face Clustering

Clusters face observations across the timeline to identify both distinct speakers and their median positions.

Smart Validation

Distinct Panel Verification

Validates that both panels show different individuals rather than duplicating the same person top and bottom.

Safe Alignment

Unified Caption Positioning

Draws captions directly over the speaker panel without colliding with either face or platform UI elements.

Split Gameplay

Gameplay & B-Roll Support

Allows replacing the lower panel with engaging gameplay footage or B-roll cutaways to boost viewer retention.

Technical Benchmark

Single crop vs stacked split screen for interviews

Keep both the speaker and the listener's reactions in every shot.

Feature

ClipNepal

Other Tools

Guest Reactions

✓Both host and guest remain visible simultaneously

✕Guest is completely cut out while the host is speaking

Wide Angle 16:9 Interviews

✓Extracts two separate 1:1 or 4:3 panels and stacks them vertically

✕Letterboxes the wide shot with massive black bars top and bottom

Caption Placement

✓Calibrated to sit comfortably within the active conversation zone

✕Overlays directly across the lower speaker's mouth or chin

Setup Complexity

✓100% automated detection; zero manual masks or crop keyframes

✕Requires 15 minutes of manual duplicate layer masking in Premiere or CapCut

Simple 3-Step Flow

How to create clips in 3 minutes

Upload your two-person interview

Submit any podcast or dialogue recorded with two or more participants.

Enable Split Layout

ClipNepal automatically identifies two distinct people and creates a stacked 2-up composition.

Export ready for social

Review both panels, customize subtitles, and download the finished vertical video.

Frequently Asked Questions

Questions creators ask

What if there are 3 or 4 speakers?

ClipNepal supports three-person stacked layouts and four-person 2x2 grid layouts, configurable in your project render preferences.

Can I switch back to single-speaker focus?

Yes. In the Clip Editor, you can toggle between 'Auto Layout', 'Split Layout', and 'Single Speaker Fill' with a single click.

Ready to automate your clip production?

Join Nepali podcasters, video creators, and agencies saving 10+ hours every week.

Start clipping free