How Klamra works

Klamra turns raw talking-head footage into a polished video or podcast — 100% locally on your Mac, powered by on-device transcription and speech AI.

Create your first project →Explore system workflow

From raw media to finished export

Four simple stages form a repeatable editing pipeline.

Import

Klamra creates local proxies and extracts high-resolution audio.

Analyze

On-device Whisper generates transcripts; pause and bad-take detection runs.

Edit

Cuts appear synchronized across text and the magnetic timeline.

Export

Finished video exports to MP4 (H.264/HEVC) or clean M4A audio.

Original files are never overwritten

Klamra preserves all source media completely untouched. Every cut is fully reversible, and you can always toggle between the edited preview and original source.

Non-destructive text and timeline edits
All video and audio processed locally on Apple Silicon
On-device AI models run offline after initial download

Documentation by topic

EDITINGTranscription & Cuts

From automatic suggestions to surgical cuts.

→AUDIOPure Voice

Noise reduction, vocal profiles, and A/B checks.

→SYSTEMArchitecture

Pipelines, local security, and workload queue.

→