AI Auto Captions

AI auto captions
transcribe, time, and burn locally

Generate accurate captions automatically from local transcription. EditAssist transcribes every word on your own machine, times them precisely, and burns them into the video or exports as SRT/VTT. No cloud upload, no per-minute fees.

01

Workflow

Ingest your video and the agent transcribes every word locally using Whisper-based models. It detects speaker changes, handles overlapping dialogue, and supports multiple languages. The resulting captions are timed to the millisecond and can be reviewed and edited as text before export.

Burn captions directly into the video with customisable fonts, positions, and safe-area margins, or export as SRT, VTT, or embedded track for broadcast compliance. The entire pipeline runs on your GPU. No per-minute fees, no upload limits, no cloud dependency.

02

Prompts

  • Transcribe the French interview and translate the subtitles to English SRT.

  • Create closed captions for ADA compliance on every export.

  • Burn subtitles into the screener copy and add the production company logo as a bug.

  • Export a clean version and a subtitled version of the final cut for the distributor.

  • Transcribe all three languages in this interview and export separate caption files.

  • Add chapter markers wherever I say 'next step' in the tutorial voice-over.

  • Create chapter markers at each topic change in the interview and export for YouTube.

  • Generate a paper edit from the transcript: keep only the answers about funding.

03

Features

Local Whisper transcription

Runs on your own GPU. No audio uploaded. Supports multiple languages, speaker detection, and overlapping dialogue.

Precise timing

Captions timed to the millisecond with word-level alignment for accurate burn-in and readable pacing.

Burn-in & export

Burn directly into video with custom styling, or export as SRT, VTT, or embedded broadcast track.

No per-minute fees

Local models are free. No upload limits, no cloud dependency, no usage metering on transcription.

04

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.

05

Getting started

  1. 01Download EditAssist for macOS or Windows and create an account. You get £15 credit and no card is needed.
  2. 02Ingest your footage. The agent transcribes, indexes, and clusters automatically.
  3. 03Describe what you want in plain English and let the agent build the cut.
  4. 04Review, refine with conversational commands, and export to your delivery spec.

More tools: AI transcript editor, Speaker diarization tool, YouTube Shorts generator. Or see everything EditAssist connects to.