Free: 3 credits (10 minutes each); Starter: $9 for 12 credits (1 hour each); Clipper: $19 for 30 credits (1 hour each); Pro: $49 for 80 credits (1 hour each); Export unlock: $2.99 per video. One credit clips one hour of the stretch you pick. Credits never expire.
Key features
Pyramid-JiT (P-JiT) architecture with 32×32 token reduction
4× more aggressive token reduction than Linum v2
Converges in 11× fewer samples than Linum v2
2B-parameter model with frozen Qwen text encoder
Pixel-space diffusion with multi-resolution readouts
PixelREPA alignment with DINOv3-L
Curriculum learning over noise schedule
User selects which clips to keep before rendering
Correct caption text before burning in
Auto-detect, Facecam, Split and Talking head layouts
Word-timed captions burned into clips
Support for English, Hindi and Hinglish captions
No watermark on any plan
Paste YouTube, Twitch, Google Drive links or upload files
What makes it different
Low-cost training: full 2B model trained for just $21K in compute
Controls baked into model training rather than post-hoc fine-tuning
Pixel-space compression without separate VAE bottleneck
Apache 2.0 open release with full model weights and code
Bill for the stretch you pick, not every minute you upload
One-time credit packs instead of monthly subscriptions
User controls which clips are made—AI only finds moments
Clip videos you don't own or create
Free plan with 3 credits and no credit card required