Welcome to Captyne
Captyne generates fully animated, word-by-word captions directly inside After Effects. No round-tripping, no manual keyframing. Here is everything you need to know.
Requirements
- Platform: Windows 10/11 (64-bit) or macOS 12+ (Apple Silicon only)
- After Effects: CC 2019 through 2026
- RAM: 8 GB minimum, 16 GB recommended
- Disk Space: ~4 GB total (3 GB Whisper model + ~200 MB runtime)
- Internet: Required for activation and cloud providers. Offline transcription runs 100% locally.
- GPU (offline transcription only): Windows needs an NVIDIA RTX 20-series / GTX 16-series or newer; Macs use Apple Silicon. Cloud providers need no GPU.
The Whisper Large v3.1 model (~3 GB) downloads separately on first use of offline transcription. The plugin prompts you automatically if it's missing.
The Captyne Workflow
Captyne has three tabs: Setup, Preview, and Account. The main workflow lives across Setup and Preview.
Step 1 — Setup & Transcribe
Everything you configure before transcribing lives in the Setup tab, top to bottom.
Reveal Mode
Controls how words are revealed in the AE timeline:
- Group All At Once — every word in a phrase appears together
- Cumulative Words — words reveal one by one within each phrase
- Cumulative Lines — each line reveals as a whole unit
- Single Word — one word at a time, each replacing the last
Transcription Providers
- Offline / Local (Free) — Whisper Large v3.1 on your own machine. Private, no usage cost, works without internet after model download.
- OpenAI — Whisper-1 — OpenAI's cloud Whisper. Requires your own API key.
- Groq — Whisper-large-v3 — Fast cloud Whisper via Groq. Requires your own API key.
- AssemblyAI — Universal-2 — High-accuracy cloud model. Requires your own API key.
- Managed API — Deepgram Nova-3 ✦ — Captyne-hosted, no key needed. Uses your plan's monthly quota.
API keys for OpenAI, Groq, and AssemblyAI are entered in Settings (gear icon inside the plugin). The Managed API option is available on paid plans and requires no configuration.
Step 2 — Preview & Edit
After transcription, Captyne automatically switches to the Preview tab. This is where you review word groupings and make edits before generating.
- Word chips — each word appears as a chip. Groups are separated visually. The timing badge on each group shows its duration.
- Merge words — hover between two words in the same group and click the chain icon
⛓to combine them into one unit (hidden while Active Highlight is on and in Single Word mode) - Split a group — hover between groups and click
✂to enter split mode, then click the word gap where you want the split - Move a word to another line — use the
↑/↓chevron buttons on each word chip to move it to the line above or below - Set word role — click a chip to cycle Normal → Emphasis → Style 3, or select it and press a number key. The chip colour updates to match.
- Edit word text — double-click a chip to correct a transcription error inline
- Delete a word — hover a chip and click the red
×; its time goes to the neighbouring word so captions stay in sync - Word Grouping toggle — switch between Auto grouping and Manual grouping using the toolbar toggle
- All Caps toggle — enables ALL CAPS output without editing individual words
Press Esc or click Cancel at any time to exit split-pick mode without making changes.
Your edits are saved automatically. Captyne stores the transcript and every Preview edit per composition in a .captyne folder next to your project and restores them when you reopen it. Use Transcript & Data → ↑ Export Session / ↓ Import Session to move captions to another project. Re-transcribing the same audio with the same settings reuses the earlier result and uses no minutes.
Step 3 — Generate Captions
When the preview looks right, click Generate Captions at the bottom of the Preview tab. Captyne will:
- Build the null structure in your AE comp: a master null plus one null per phrase
- Create one text layer per word with correct timing
- Apply your selected preset's effects and animations to each word
- Place AE markers for every word so you can drag to adjust timing
Run Generate Captions again at any time to re-generate with updated settings. Previous caption layers are replaced automatically.
Presets
Each preset defines fonts, effects, animations, layout, and style roles (Normal, Emphasis, Style3). They appear in the top card of the Setup tab.
Creating Your Own
Click + New Preset in the My Presets section. Captyne creates a tagged template comp in AE — design your caption style freely inside it, then click ⊙ Capture Preset in the plugin to save it.
While a preset is open, the AE Style Layer Tools let you tag style layers, add companion layers (box, text copy, burst, arrow, underline, circle scribble, sparkle) and add Smart Effects — Shine, Gradient Reveal, Glow Pulse, Blur → Focus, Slide / Pop / Spin / Bounce transitions, Glitch and Partial Blur. They track each word’s size and position and are timed by markers. See the word-tracking preset guide.
Use ↑ Export / ↓ Import to share presets as .json files with other users.
Installed presets are stored in:
Windows: %AppData%\Adobe\CEP\extensions\Captyne\presets\
macOS: ~/Library/Application Support/Adobe/CEP/extensions/Captyne/presets/
Presets that use third-party plugins (Deep Glow, Optical Flares, etc.) show a warning badge if that plugin isn't installed on your machine. Built-in AE effects always work.
Emphasis Strategy
The Emphasis Strategy card in the Setup tab controls how words are assigned the Emphasis style role before generating.
- Default (Last Word) — the last word in each group automatically gets the Emphasis role. Good for most caption styles.
- AI Decide — the Managed API picks the most important words while it transcribes. Requires the Managed API (Pro plan or higher); with other providers it is locked and the preset’s default emphasis is used.
- Manual — no automatic assignment. Click word chips in the Preview tab to set emphasis yourself.
Interactive Preview
The Preview tab shows a live word-by-word view of your transcript grouped into caption phrases. Everything here feeds directly into Generate Captions.
Group Controls
Between each group there is a gap zone. Hover over it to reveal two inline buttons:
- ✂ Split — enters split-pick mode. The group gets a red border and word gaps become clickable red bars. Click any gap to split there, or press Esc to cancel.
- ⛓ Merge — merges the group above with the group below into one phrase (only available when a group below exists).
Word Controls
Each word chip has controls on hover:
- Click — cycles Normal → Emphasis → Style 3. Chip colour matches the role.
- Double-click — edit the word’s spelling
- Red × — delete the word; its time goes to the neighbouring word
- ↑ / ↓ chevrons — move the word to the line above or below within its group
- ⛓ between chips — hover between two words and click to merge them into one chip
- ✕ on merged chips — unmerge a previously merged word pair
Active Highlight
Active Highlight, in the Preview tab, emphasises each word while it is being spoken — karaoke-style captions. It is built into the caption layers when you click Generate Captions.
Enable it with the toggle, then configure:
- Font colour — colour of the active word during playback
- Stroke — optional outline on the active word, with configurable width and colour
- Highlight box — optional coloured box behind the active word, with padding, rounded corners and a shadow
- Drop shadow and a short scale pop on the active word
Active Highlight settings are marked Runtime Only because they belong to your current captions and are not saved into the preset. Active Highlight has no effect in Single Word reveal mode, and word merging is hidden while it is on.
Caption Layer Structure
Every generation creates a structured set of layers in your AE composition:
- ACP_MASTER — move, scale or rotate to reposition all captions at once
- ACP_SET_n_CTRL — one null per caption phrase. Move it to shift that phrase, for example out of the way of a face or graphic.
- ACP_SET_n_WORD_m — one text layer per word. Effect points such as gradients and light sweeps follow each word wherever you move it.
Marker-Driven Timing
Captyne places AE markers on every word layer. Each word's in/out animation is driven by those markers — drag one to shift timing without touching keyframes.
- In the AE timeline, expand a word layer and find its marker
- Drag the marker left or right to shift when that word animates in or out
- All animations follow automatically — no keyframe editing needed
Every word has an IN_END marker (in-animation finished) and an OUT_START marker (exit begins). Presets with a sweep, glow or blur also add SE_IN and SE_OUT for that effect. Building your own? See Tutorial 09.
Offline AI Transcription
The Offline / Local provider runs Whisper Large v3.1 entirely on your machine — your footage never leaves your computer. On Windows it needs an NVIDIA RTX 20-series / GTX 16-series GPU or newer; on Mac it runs on Apple Silicon. On unsupported PCs, Captyne turns the option off and you can use a cloud provider instead.
Downloading the Model
If the model isn't present, the plugin shows a Whisper Model Not Found card at the top of the Setup tab with three options:
- ⬇ Download Now (~3 GB) — downloads directly from inside the plugin with a progress bar
- 📋 Copy Download Link — copies the URL so you can download with your browser or a download manager
- On Mac, double-click the Download Captyne Model shortcut the installer puts on your Desktop
The model is saved to:
- Windows:
C:\Captyne\models\ggml-large-v3.1.bin - macOS:
~/CaptyneRuntime/models/ggml-large-v3.1.bin
Performance Tips
- Keep your NVIDIA driver up to date on Windows
- Older cards such as the GTX 10-series can’t run offline transcription — use the Managed API instead
- Apple Silicon GPU is used automatically on Mac (Metal acceleration)
- Trim your comp to just the dialogue section for the fastest results
Plugin Tabs Reference
- Presets — My Presets and Marketplace sub-tabs; select, create, export, import
- Emphasis Strategy — Default (Last Word), AI Decide, or Manual
- Source Type — Active AE Comp or Manual Media File
- Provider — Offline, OpenAI, Groq, AssemblyAI, or Managed API
- Language — Auto-detect or manual selection (99 languages)
- Speakers — Off or On for multi-speaker detection
- Model — Default / Best / Fast quality hint for cloud providers
- Reveal Mode — Group All At Once / Cumulative Words / Cumulative Lines / Single Word
- Words / Phrase — max words per caption phrase
- Words / Line — max words per line within a phrase
- Pause threshold — silence gap (seconds) that triggers a new phrase
- Transcribe button — runs transcription and loads results into the Preview tab
- Interactive Preview — word chips grouped into phrases; merge, split, move, role-assign
- Word Grouping toggle — Auto or Manual grouping mode
- All Caps toggle — ALL CAPS output without editing individual words
- Active Highlight — configure runtime word highlight (font colour, stroke, bounding box)
- Generate Captions button — builds animated caption layers in your AE comp
- Sign In — enter your captyne.com email and password and click Sign In & Activate
- Your Account — shows your name, plan and status, with Sign Out
- Settings — gear icon opens provider API keys, offline runtime path, and other preferences
FAQ
Captyne doesn't appear in Window → Extensions
Restart After Effects. If it still doesn't appear, go to Edit → Preferences → Scripting & Expressions and enable Allow Scripts to Write Files and Access Network.
Transcription fails or produces no output
If using Offline, check that the model is downloaded (the plugin shows a card if it's missing). Also ensure your comp has audio — Captyne won't transcribe silent compositions.
My preset shows a warning badge
The preset requires a third-party plugin or font not installed on your machine. Install the missing item and try again, or choose a preset that uses only built-in AE effects.
Can I edit captions after generating?
Yes. Word layers are standard AE text layers. Edit text directly in AE, drag markers to change timing, or go back to the Preview tab and click Generate Captions again.
How do I sign in?
Open the plugin, go to the Account tab, enter your captyne.com email and password, and click Sign In & Activate. Beta testers use the password they set from their approval email. Forgot it? Reset your password.
It says my device limit is reached
Each plan allows a set number of devices. Deactivate an old device in your account at captyne.com/account, then sign in again. Still stuck? Contact [email protected].
Will I lose my captions if I close the project?
No. Captyne saves the transcript and your Preview edits next to the project and restores them when you reopen it.
Support
Found a bug or need help? Please reach out — your feedback directly shapes Captyne.
- Email: [email protected]
- Include your AE version, OS version, and a description of what happened
- Windows crashes: attach the AE log from
%AppData%\Adobe\After Effects\[version]\Logs\ - Mac crashes: attach the AE log from
~/Library/Logs/Adobe/After Effects/