Every tool your shorts operation needs
A complete pipeline from owned long-form content to publishable, monetization-ready clips — with human review built in at every step.
Paste a YouTube URL (or upload a file) and VolfPack handles the rest: download with yt-dlp, crop landscape to 9:16, split at natural segment boundaries, clean frames, transcribe with Whisper, rewrite for shorts pacing, generate TTS audio, burn word-level captions, and assemble the final export — all in one durable async job.
Create a channel for each YouTube presence and lock in its crop strategy, voice, language, caption style, and LLM rewrite prompts. Every new project inherits those settings instantly — no per-clip setup.
Generate natural-sounding TTS with ElevenLabs (hosted) or Bark (local GPU). Word-level timestamps are aligned automatically so captions snap to the exact spoken word. Style them per channel — colour, size, highlight timing.
Every AI step — STT, LLM, TTS, OCR — runs through a swappable adapter. Use OpenAI today, switch to a local Whisper or LiteLLM-compatible model tomorrow. Bring your own API keys or use managed credits; the pipeline stays the same.
Four dedicated editors — crop, split timeline, script, captions — let you correct AI output before it moves to the next phase. Nothing exports without your approval. Retry individual phases without re-running the whole pipeline.
Import views, retention, and revenue data per exported short. Over time, VolfPack learns which hooks, caption styles, and segment types drive the best results on each channel and surfaces those patterns as recommended templates.
Run the full stack locally on a production PC via Docker Compose — your data and API keys never leave the machine. When your team grows, switch to the hosted tier for managed storage, a shared review queue, and team roles — same feature set, no migration.