ScriptFollow listens to actors on stage and auto-scrolls the script in real time. Mark cues anywhere, fire them on a key press, and never lose your place again.
Drop in any script — Fountain or PDF — and ScriptFollow does the rest. Cues, notes, and live tracking are built right in. No imports, no exports, no fuss.
One window. The script you know, with calling intelligence woven through it.
ScriptFollow uses Double Metaphone phonetic matching, not literal text. Dropped articles, swapped words, and stumbles don't break the follow — the cursor stays anchored to the next syllable that matters.
Lighting, sound, fly, deck — every cue lives on the line that triggers it. Reorder, renumber, and edit during tech without leaving the page. Auto-saved per script.
Every action is one keystroke away. The interface stays out of the way, the script stays large, and your hands never leave the home row.
Drop a PDF and ScriptFollow uses Claude or GPT to extract speakers and lines into a clean structure. Review, tweak, and load. Fountain works too, of course — natively.
Call your show without burying your nose in the prompt book. Eyes up, on the actors and the deck — where they belong.
Track blocking, line notes, and pickups in one place. Keep a clean record across every rehearsal and run.
Sit next to the SM, follow the same script, and place provisional cues without disturbing the call.
Yes — Fountain (.fountain, .txt) loads natively, and PDFs are parsed by an AI provider of your choice (Claude Haiku or GPT-4o mini). Once parsed, the structured script is cached locally so you only do it once.
The follower uses phonetic matching with a forward-and-backward search window, so it tolerates word swaps, dropped articles, and stumbles. If the cast jumps ahead by a page, it usually catches up within a few words. You can also tap SPACE to nudge it manually at any time.
Audio is streamed to Deepgram for real-time transcription and is not stored. PDF parsing calls happen once per script and the results are cached on your machine. Bring your own API keys — nothing routes through us.
Yes for Fountain scripts and manual line tracking. The auto-follow listening feature requires an internet connection for transcription. PDF parsing requires it once, then runs from cache.
The app is free during the beta. You provide your own Deepgram and Anthropic/OpenAI API keys, which are typically a few cents per show.
macOS only for now (Apple Silicon and Intel). Windows is on the roadmap. iPad is a longer-term ambition once the core feels right on desktop.
Free during beta. Built by a stage manager who got tired of losing his place.