Rough-cut from transcript
edit the text, build the timeline
Generate rough cuts directly from transcripts, scripts, and paper edits. EditAssist transcribes every word locally, then builds timeline assemblies from your text selections, scene descriptions, and narrative notes. Frame-accurate and ready for editorial.
Workflow
Upload your transcript, script, or paper edit and the agent parses the narrative structure. It matches scene descriptions to the best takes, builds an assembly timeline with synced audio, and adds scene markers. You refine by editing the text; changes propagate to the timeline in real time.
The agent understands speaker labels, overlapping dialogue, and multi-language transcripts. It can build thematic assemblies, chronological string-outs, or narrative restructures from plain-English instructions. All transcription and assembly runs on your own machine.
Prompts
“Generate a paper edit from the transcript: keep only the answers about funding.”
“Build an assembly from the shooting script: match each scene description to the best take.”
“Build a narrative structure from the interviews: group by theme, not chronology.”
“Build a paper edit of the 3-hour assembly: keep only the strongest narrative beats.”
“Transcribe all 60 interviews, identify the 12 key characters, and build a selects reel per person.”
“Generate a story map, every scene, every character, every location, from the raw transcripts.”
“Find the moment where she says 'that changed everything' and mark it as selects.”
“Chapter this tutorial into sections automatically.”
Features
Script-to-timeline assembly
Parses scripts, transcripts, and paper edits to build frame-accurate assemblies with the best take per scene.
Text-driven refinement
Edit the transcript and changes propagate to the timeline. Delete, reorder, or select text to reshape the cut.
Thematic & narrative modes
Build chronological string-outs, thematic groups, or narrative arcs from plain-English instructions.
Local transcription
The transcription engine runs on your own GPU. No audio uploaded. Multi-language and speaker-label support.
Local-first
All transcription, vision analysis, semantic search, and clustering run on your own GPU. Nothing about your footage is uploaded. The only data that leaves is the text of your conversation with the agent. It's free to download, local models are free, and a new account starts with £15 of credit and no card. For long-running cloud work, pay as you go, or choose £20/mo Starter or £100/mo Pro.
FAQ
How accurate is the mapping from transcript text to timeline cuts?
Every text edit maps to precise timecode, so the cut stays frame-accurate. Select a sentence in the transcript and the corresponding clip range is trimmed on the timeline exactly; rearrange paragraphs and the timeline reorders to match. The transcript and the timeline never drift apart.
Can I build a rough cut from a transcript without exact keywords in the footage?
Yes. Search over the transcript is semantic: you can ask for every mention of a topic or every moment a theme comes up, not just literal keyword matches. The agent finds the segments by meaning, marks them as selects, and builds the assembly from those selections.
Does the transcript-based rough cut handle filler words and silences?
The agent detects ums, ahs, breaths, and dead air from the transcript and waveform, and can tighten them out of the assembly with clean crossfades as part of the build, so you start from a cut that already flows.
Getting started
- 01Download EditAssist for macOS or Windows and create an account. You get £15 credit and no card is needed.
- 02Ingest your footage. The agent transcribes, indexes, and clusters automatically.
- 03Describe what you want in plain English and let the agent build the cut.
- 04Review, refine with conversational commands, and export to your delivery spec.
More tools: AI transcript editor, Rough cut generator, Speaker diarization tool. Or see everything EditAssist connects to.