Installed in two minutes.

Requirements

• Adobe Premiere Pro 2020 (v14) or newer, Windows or macOS.
• An API key from OpenAI (platform.openai.com/api-keys) or Groq (console.groq.com/keys- Groq's whisper-large-v3 is fast and currently very cheap).
• Internet access while transcribing - rendering and placement are fully local.
• Optional: ffmpegon your PATH lets the panel auto-compress files bigger than the API's 25 MB upload cap.

Install

Windows

Right-click install/install-win.bat → run it. Admin rights are NOT needed - it only writes per-user settings. Restart Premiere.

macOS

Open Terminal and run:

chmod +x install/install-mac.sh && ./install/install-mac.sh

Restart Premiere, then open Window ▸ Extensions ▸ Caption Plug.

The installer flips Adobe's PlayerDebugModeflag so Premiere will load self-installed panels - the standard way to run CEP extensions you didn't get from the marketplace. It also copies the 62 bundled caption fonts into your per-user font folder (no admin rights needed) so they're usable in Essential Graphics; the panel's own rendering works even if that step is skipped.

Use it

The panel is organised into collapsible sections (Animation preset, Style, SFW mode, Transcription, Output, Storage) - closed sections show their current setting on the right, and the preview plus the Generate button stay visible at all times. First time? The panel opens a guided tutorial on first launch (reopen it any time with the ? Tutorial button in the top bar).

1. Open the sequence you want captioned - the status pill in the panel's header shows the sequence it's locked onto, with its resolution and fps.
2. Pick an animation preset and tune the style. The live preview at the top runs the actual render engine, so what you see is literally what gets rendered. The color wells open a built-in picker with a saturation/hue surface, hex field and preset swatches.
3. Optional: open SFW mode to censor inappropriate words and add bleeps - see the censor demo.
4. Under Transcription, choose a provider, paste your API key (saved locally), and give it audio: select your dialogue clips in the timeline and hit Use selected clips. No setup needed - the panel finds Premiere's own MP3 encoder preset, mutes unselected audio like music beds during the export, and offsets the captions so they land exactly where those clips sit. With nothing selected it exports the whole sequence. Or use Choose audio / video file… to point it at a file directly.
5. Pick an Output mode - Animated clips (rendered PNGs, full animation) or Editable text (native caption track) - and hit Generate. Rendered caption media is saved into a Caption Plug Media folder next to your project file, so the project relinks fine later.

Want a plain SRT file on disk (for YouTube, other editors…)? Export .srt only transcribes and saves a standard SRT wherever you choose, and also attaches it as a caption track on Premiere versions that support it.

The two output modes

• Animated clips- transparent PNG image-sequence clips rendered at your sequence's exact resolution and frame rate, placed on a fresh video track. Because every frame is rasterized against the sequence fps, sync is frame-accurate by construction - no drifting keyframes, no manual nudging. Frames where nothing moves are stored as hard links rather than copies, so disk use stays far below one-PNG-per-frame.
• Editable text- a native Premiere caption track instead of rendered media: the text stays editable in the caption editor, styling lives in Essential Graphics (set it once, save a Track Style), and almost nothing touches the disk. The per-word animation presets don't apply in this mode - that's the trade.

Troubleshooting

The panel doesn't appear under Window ▸ Extensions

PlayerDebugModedidn't stick - rerun the installer, make sure Premiere was fully closed, and on macOS run killall cfprefsd afterwards so the preference cache reloads.

"File is over the 25 MB API limit"

Whisper's upload cap. Export your audio as MP3 (File ▸ Export ▸ Media ▸ Format: MP3) and pick that file, or install ffmpeg and the panel compresses automatically.

Captions import but look short/long

The panel conforms still-sequences to your fps and trims each clip, but if you're on a very old Premiere build, set Preferences ▸ Media ▸ Indeterminate Media Timebaseto your sequence's frame rate before generating.

Wrong words / names misspelled

That's the transcription model. Groq's whisper-large-v3 is noticeably better than whisper-1 on names; you can also fix a word by re-rendering after editing - or use Export .srt onlyand edit text in Premiere's caption editor instead.

Premiere freezes for a moment on "Use selected clips"

Normal - Premiere's export call is blocking. The render/transcribe steps afterwards don't block the app.

"Couldn't find an audio encoder preset"

Rare - the panel reads the MP3 presets inside your Premiere install, and some IT-managed installs strip them. Save one yourself once; the plugin's epr/README.txt explains how.

Captions from selected clips land at the wrong time

Make sure the clips were still selected when you clicked the button; the exported audio is stamped with the selection's start time and the captions are offset to match it. Re-export after moving clips.