AI transcript editor
cut from text, not from timecode
Edit video by editing the transcript. EditAssist transcribes every word locally, then lets you cut, rearrange, and refine by selecting text. Changes propagate back to the timeline automatically. Works in DaVinci Resolve, Premiere Pro, After Effects and Final Cut.
Workflow
Ingest your footage and the agent transcribes every speaker locally using Whisper-based models. The transcript becomes your editing interface: select sentences, delete ums and ahs, rearrange paragraphs, and the timeline updates in real time. Every text edit maps to precise timecode so the cut stays frame-accurate.
You can search the transcript for specific phrases, find every mention of a topic, or build a paper edit from a theme. The agent understands speaker labels, overlapping dialogue, and multiple languages. Export the refined transcript as SRT, VTT, or burn it directly into the video.
Prompts
“Find the moment where she says 'that changed everything' and mark it as selects.”
“Transcribe all dialogue in the assembly and find every mention of the suspect's name.”
“Generate a paper edit from the transcript: keep only the answers about funding.”
“Search the transcript for every question the interviewer asks and list them with timecodes.”
“Find all the moments where I say 'um' or 'uh' and remove them.”
“Detect every moment the interview subject goes off-topic.”
“Chapter this tutorial into sections automatically.”
“Build chapter markers from the slide transitions in this lecture.”
Features
Local transcription
Whisper-based models run on your own GPU. No audio is uploaded. Supports multiple languages and speaker labels.
Text-to-timeline editing
Select, delete, and rearrange text; changes propagate to the timeline with frame-accurate timecode mapping.
Semantic transcript search
Find every mention of a topic, every question, or every emotional moment across the entire transcript.
Auto-caption export
Export SRT, VTT, or burn captions directly into the video with precise timing and speaker attribution.
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.
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 auto captions, Speaker diarization tool, Natural language video editing. Or see everything EditAssist connects to.