Enhance video creation by importing, customizing, and styling captions directly within the editor.

Prerequisites#
Before you begin, make sure you have the following:
- Node.js v20+ and npm installed locally – Download Node.js
- A supported browser – Chrome 114+, Edge 114+, Firefox 115+, Safari 15.6+
See Browser Support for the full list.
Get Started#
Start fresh with a standalone Video Captions project. This creates a complete, ready-to-run application.
Step 1: Clone the Repository#
git clone https://github.com/imgly/starterkit-video-captions-react-web.gitgit clone https://github.com/imgly/starterkit-video-captions-react-web.gitnpx degit imgly/starterkit-video-captions-react-web starterkit-video-captions-react-webnpx degit imgly/starterkit-video-captions-react-web starterkit-video-captions-react-webThe src/ folder contains the editor code:
src/├── app/ # Demo application├── imgly/│ ├── config/│ │ ├── actions.ts # Export/import actions│ │ ├── features.ts # Feature toggles│ │ ├── i18n.ts # Translations│ │ ├── plugin.ts # Main configuration plugin│ │ ├── settings.ts # Engine settings│ │ └── ui/│ │ ├── canvas.ts # Canvas configuration│ │ ├── components.ts # Custom component registration│ │ ├── dock.ts # Dock layout configuration│ │ ├── index.ts # Combines UI customization exports│ │ ├── inspectorBar.ts # Inspector bar layout│ │ ├── navigationBar.ts # Navigation bar layout│ │ └── panel.ts # Panel configuration│ ├── index.ts # Editor initialization function│ └── plugins/│ └── auto-caption.ts└── index.tsx # Application entry pointStep 2: Install Dependencies#
Install the required packages:
cd starterkit-video-captions-react-web
npm installcd starterkit-video-captions-react-web
npm installcd starterkit-video-captions-react-web
pnpm installcd starterkit-video-captions-react-web
pnpm installcd starterkit-video-captions-react-web
yarncd starterkit-video-captions-react-web
yarnStep 3: Download Assets#
CE.SDK requires engine assets (fonts, icons, UI elements) to function. These must be served as static files from your project’s public/ directory.
curl -O https://cdn.img.ly/packages/imgly/cesdk-js/1.73.0/imgly-assets.zip
unzip imgly-assets.zip -d public/
rm imgly-assets.zipcurl -O https://cdn.img.ly/packages/imgly/cesdk-js/1.73.0/imgly-assets.zip
unzip imgly-assets.zip -d public/
rm imgly-assets.zipconst config = { // ... baseURL: '/assets' // ...};Step 4: Run the Development Server#
npm run devnpm run devpnpm run devpnpm run devyarn devyarn devOpen http://localhost:5173 in your browser.
Get Started#
Integrate the Video Captions into an existing SvelteKit application. This adds the editor configuration to your current project structure.
Step 1: Clone#
cd your-projectcd your-projectClone the starter kit and copy the editor configuration to your project:
git clone https://github.com/imgly/starterkit-video-captions-react-web.git
cp -r starterkit-video-captions-react-web/src/lib/imgly ./src/lib/imgly
rm -rf starterkit-video-captions-react-webgit clone https://github.com/imgly/starterkit-video-captions-react-web.git
cp -r starterkit-video-captions-react-web/src/lib/imgly ./src/lib/imgly
rm -rf starterkit-video-captions-react-webnpx degit imgly/starterkit-video-captions-react-web/src/lib/imgly ./src/lib/imglynpx degit imgly/starterkit-video-captions-react-web/src/lib/imgly ./src/lib/imglyThe imgly/ folder contains the editor configuration:
imgly/├── config/│ ├── actions.ts # Export/import actions│ ├── features.ts # Feature toggles│ ├── i18n.ts # Translations│ ├── plugin.ts # Main configuration plugin│ ├── settings.ts # Engine settings│ └── ui/│ ├── canvas.ts # Canvas configuration│ ├── components.ts # Custom component registration│ ├── dock.ts # Dock layout configuration│ ├── index.ts # Combines UI customization exports│ ├── inspectorBar.ts # Inspector bar layout│ ├── navigationBar.ts # Navigation bar layout│ └── panel.ts # Panel configuration├── index.ts # Editor initialization function└── plugins/ └── auto-caption.tsStep 2: Install Dependencies#
Install the required packages for the editor:
Core Editor#
Install the Creative Editor SDK:
npm install @cesdk/cesdk-jsnpm install @cesdk/cesdk-jspnpm add @cesdk/cesdk-jspnpm add @cesdk/cesdk-jsyarn add @cesdk/cesdk-jsyarn add @cesdk/cesdk-jsAutocaption Plugin#
Add AI-powered autocaption generation:
npm install @imgly/plugin-autocaption-webnpm install @imgly/plugin-autocaption-webpnpm add @imgly/plugin-autocaption-webpnpm add @imgly/plugin-autocaption-webyarn add @imgly/plugin-autocaption-webyarn add @imgly/plugin-autocaption-webStep 3: Download Assets#
CE.SDK requires engine assets (fonts, icons, UI elements) to function. For SvelteKit projects, place these in your static/ directory which is served automatically.
curl -O https://cdn.img.ly/packages/imgly/cesdk-js/1.73.0/imgly-assets.zip
unzip imgly-assets.zip -d static/
rm imgly-assets.zipcurl -O https://cdn.img.ly/packages/imgly/cesdk-js/1.73.0/imgly-assets.zip
unzip imgly-assets.zip -d static/
rm imgly-assets.zipStep 4: Create the Editor Component#
Choose your editing mode and initialize the editor accordingly:
<script> import { onMount, onDestroy } from 'svelte'; import CreativeEditorSDK from '@cesdk/cesdk-js'; import { initVideoCaptionsAutocaptionEditor } from './imgly';
let container; let cesdk = null;
onMount(async () => { cesdk = await CreativeEditorSDK.create(container, { baseURL: '/assets' }); await initVideoCaptionsAutocaptionEditor(cesdk); });
onDestroy(() => { if (cesdk) { cesdk.dispose(); } });</script>
<div bind:this={container} style="width: 100vw; height: 100vh;"></div><script> import { onMount, onDestroy } from 'svelte'; import CreativeEditorSDK from '@cesdk/cesdk-js'; import { initVideoCaptionsBlankEditor } from './imgly';
let container; let cesdk = null;
onMount(async () => { cesdk = await CreativeEditorSDK.create(container, { baseURL: '/assets' }); await initVideoCaptionsBlankEditor(cesdk); });
onDestroy(() => { if (cesdk) { cesdk.dispose(); } });</script>
<div bind:this={container} style="width: 100vw; height: 100vh;"></div><script> import { onMount, onDestroy } from 'svelte'; import CreativeEditorSDK from '@cesdk/cesdk-js'; import { initVideoCaptionsImportEditor } from './imgly';
let container; let cesdk = null;
onMount(async () => { cesdk = await CreativeEditorSDK.create(container, { baseURL: '/assets' }); await initVideoCaptionsImportEditor(cesdk); });
onDestroy(() => { if (cesdk) { cesdk.dispose(); } });</script>
<div bind:this={container} style="width: 100vw; height: 100vh;"></div><script> import { onMount, onDestroy } from 'svelte'; import CreativeEditorSDK from '@cesdk/cesdk-js'; import { initVideoCaptionsPreCaptionedEditor } from './imgly';
let container; let cesdk = null;
onMount(async () => { cesdk = await CreativeEditorSDK.create(container, { baseURL: '/assets' }); await initVideoCaptionsPreCaptionedEditor(cesdk); });
onDestroy(() => { if (cesdk) { cesdk.dispose(); } });</script>
<div bind:this={container} style="width: 100vw; height: 100vh;"></div>Caption Modes#
The Video Captions starter kit supports four distinct workflows for working with video captions:
Autocaption Mode#
Auto-generate captions using AI-powered speech-to-text transcription with ElevenLabs Scribe V2:
await initVideoCaptionsAutocaptionEditor(cesdk);This mode:
- Loads a video with audio ready for transcription
- Integrates the
@imgly/plugin-autocaption-webplugin - Configures the ElevenLabs Scribe V2 provider via fal.ai proxy
- Automatically transcribes speech and generates styled captions
- Opens the caption inspector panel for editing
Prerequisites:
- fal.ai API key configured in
VITE_AUTOCAPTION_PROXY_URLenvironment variable - Default proxy:
https://proxy.img.ly/api/proxy/falai
Blank Mode#
Start with an empty video canvas for manual caption creation:
await initVideoCaptionsBlankEditor(cesdk);This mode:
- Creates a new blank video scene (1280x720)
- Lets you import your own videos
- Manually create and style captions from scratch
- No AI processing required
Import Mode#
Import existing SRT subtitle files and edit them:
await initVideoCaptionsImportEditor(cesdk);This mode:
- Loads a video for caption placement
- Allows importing SRT files with existing captions
- Edit timing, styling, and positioning
- Export back to video with embedded captions
Pre-captioned Mode#
Edit videos that already have embedded captions:
await initVideoCaptionsPreCaptionedEditor(cesdk);This mode:
- Loads a video with captions already embedded
- Automatically selects the first caption for editing
- Full access to caption styling and timing controls
- Refine existing captions without re-transcribing
Autocaption Setup#
To enable AI-powered caption generation, configure the fal.ai proxy URL:
import AutocaptionPlugin from '@imgly/plugin-autocaption-web';import { ElevenLabsScribeV2 } from '@imgly/plugin-autocaption-web/fal-ai';
// Configure the proxy URL (default: https://proxy.img.ly/api/proxy/falai)const AUTOCAPTION_PROXY_URL = import.meta.env.VITE_AUTOCAPTION_PROXY_URL || 'https://proxy.img.ly/api/proxy/falai';
// Add the autocaption plugin with ElevenLabs Scribe V2 providerawait cesdk.addPlugin( AutocaptionPlugin({ provider: ElevenLabsScribeV2({ proxyUrl: AUTOCAPTION_PROXY_URL }) }));The proxy URL should point to a server that forwards requests to fal.ai with your API key. For production use, deploy your own proxy to keep the API key secure.
Caption Inspector Panel#
The caption inspector panel provides comprehensive controls for editing captions:
// Open the caption inspector panel programmaticallycesdk.ui.openPanel('//ly.img.panel/inspector/caption');The panel includes:
- Caption text editor - Edit caption content with live preview
- Timing controls - Adjust start time and duration
- Styling options - Font, size, color, background, alignment
- Positioning - Move captions anywhere on the video
- Caption presets - Apply pre-designed caption styles
The panel automatically opens when using autocaption mode, but you can open it manually in any mode using the code above.
Set Up a Scene#
Each caption mode has a dedicated init function that handles scene setup automatically.
For custom workflows beyond the predefined modes, you can load content manually:
// After initialization, load your own videoawait cesdk.createFromVideo('https://example.com/video.mp4');
// Or load from a template archive with captionsawait cesdk.loadFromArchiveURL('https://example.com/template.zip');
// Or load from a scene fileawait cesdk.loadFromURL('https://example.com/scene.json');The predefined init functions load appropriate scenes:
- autocaption - Loads a video ready for AI transcription
- blank - Creates an empty 1280x720 video scene
- import - Loads a video for SRT import
- pre-captioned - Loads a video with embedded captions
Customize Assets#
The Video Captions uses asset source plugins to provide built-in libraries for video clips, audio, effects, stickers, and fonts. The starter kit includes a curated selection—customize what’s included based on your needs.
Asset sources are added via plugins in src/index.ts. Enable or disable individual sources:
import { FiltersAssetSource, StickerAssetSource, TextAssetSource, VectorShapeAssetSource, EffectsAssetSource, // ...} from '@cesdk/cesdk-js/plugins';
// Add only the sources you needawait cesdk.addPlugin(new FiltersAssetSource());await cesdk.addPlugin(new StickerAssetSource());await cesdk.addPlugin(new TextAssetSource());await cesdk.addPlugin(new VectorShapeAssetSource());await cesdk.addPlugin(new EffectsAssetSource());// ...For production deployments, self-hosting assets is required—the IMG.LY CDN is intended for development only. See Serve Assets for downloading assets, configuring baseURL, and excluding unused sources to optimize load times.
Configure Actions#
Actions are functions that handle user interactions like exporting videos, saving scenes, and importing files. CE.SDK provides built-in actions that you can run directly or override with custom implementations.
Key built-in actions:
exportDesign– Export the current video to MP4 formatsaveScene– Save the scene as a JSON string for later editingimportScene– Import a previously saved scene (supports.sceneand.cesdkformats)exportScene– Export the scene as a JSON file or.cesdkarchive with all assetsuploadFile– Handle file uploads with progress tracking
Use cesdk.actions.run() to execute any action:
// Run a built-in actionawait cesdk.actions.run('exportDesign', { mimeType: 'video/mp4' });Import from File Picker#
// Let users open videos from their devicecesdk.actions.register('importVideo', async () => { const blobURL = await cesdk.utils.loadFile({ accept: 'video/*', returnType: 'objectURL' }); await cesdk.createFromVideo(blobURL);});Export and Save#
// Register export action that downloads the edited videocesdk.actions.register('exportDesign', async (exportOptions) => { const { blobs, options } = await cesdk.utils.export(exportOptions); await cesdk.utils.downloadFile(blobs[0], options.mimeType);});Upload to Your Backend#
// Override the built-in exportDesign action to send to your servercesdk.actions.register('exportDesign', async (exportOptions) => { const { blobs } = await cesdk.utils.export(exportOptions);
const formData = new FormData(); formData.append('video', blobs[0], 'edited-video.mp4');
const response = await fetch('/api/upload', { method: 'POST', body: formData });
const { url } = await response.json(); console.log('Uploaded to:', url);});Customize (Optional)#
Theming#
CE.SDK supports light and dark themes out of the box, plus automatic system preference detection. Switch between themes programmatically:
// 'light' | 'dark' | 'system' | (() => 'light' | 'dark')cesdk.ui.setTheme('dark');See Theming for custom color schemes, CSS variables, and advanced styling options.
Localization#
Customize UI labels and add support for multiple languages. The i18n system supports translation keys for all UI elements:
// Override specific labelscesdk.i18n.setTranslations({ en: { 'actions.export.video': 'Download Video', 'common.cancel': 'Cancel', 'common.apply': 'Apply' }});
// Add a new languagecesdk.i18n.setTranslations({ de: { 'actions.export.video': 'Video herunterladen' }});
// Set the active localecesdk.i18n.setLocale('de');See Localization for supported languages, translation key reference, and right-to-left language support.
UI Layout#

Customize the editor interface by modifying the dock, inspector bar, navigation bar, and canvas menu. CE.SDK provides Order APIs to control which components appear and in what sequence.
// Get current navigation bar componentsconst navOrder = cesdk.ui.getNavigationBarOrder();
// Add a custom button to the navigation barcesdk.ui.insertNavigationBarOrderComponent( 'ly.img.spacer', { id: 'my-custom-action' }, 'after');
// Rearrange dock itemscesdk.ui.setDockOrder([ 'ly.img.assetLibrary.dock', 'ly.img.separator', 'my-custom-dock-item']);
// Customize the inspector barcesdk.ui.setInspectorBarOrder([ 'ly.img.fill.inspectorBar', 'ly.img.separator', 'ly.img.filter.inspectorBar']);The Order API methods follow a consistent pattern across all UI areas:
get*Order()– Retrieve the current component orderset*Order()– Replace the entire orderinsert*OrderComponent()– Add components relative to existing ones
See Dock, Inspector Bar, Navigation Bar, Canvas Menu, and Canvas for detailed layout customization options.
Custom Components#
Build custom UI components using the builder system and integrate them in the editor. Custom components receive reactive state updates and can interact with the engine API.
// Register a custom componentcesdk.ui.registerComponent('my-custom-button', ({ builder, engine }) => { const selectedBlocks = engine.block.findAllSelected();
builder.Button('apply-effect', { label: 'Apply Effect', isDisabled: selectedBlocks.length === 0, onClick: () => { // Apply custom logic to selected blocks } });});
// Add the component to the navigation barcesdk.ui.insertNavigationBarOrderComponent( 'ly.img.spacer', 'my-custom-button', 'after');Custom components automatically re-render when the engine state they depend on changes—no manual subscription management required.
See Register New Component for the complete builder API and component patterns.
Settings & Features#
Fine-tune editor behavior through settings and features.
Settings configure core engine behavior—rendering, input handling, and history management:
cesdk.engine.editor.setSettingBool('page/dimOutOfPageAreas', true);cesdk.engine.editor.setSettingBool('mouse/enableZoomControl', true);cesdk.engine.editor.setSettingBool('features/undoHistory', true);Features toggle which editing tools and panels appear in the UI:
// Toggle editor featurescesdk.feature.enable('ly.img.trim', true);cesdk.feature.enable('ly.img.filter', true);cesdk.feature.enable('ly.img.adjustment', true);See Settings and Features for the complete reference.
Explore Plugins#
CE.SDK has a rich plugin ecosystem that extends the editor with powerful capabilities. Plugins can add new features, integrate third-party services, or customize editor behavior.
AI Integration#
Extend the editor with generative AI capabilities for text-to-image generation, image enhancement, and intelligent editing features. CE.SDK integrates with various AI providers.
import AIPlugin from '@imgly/plugin-ai-generation';
// Configure AI generationawait cesdk.addPlugin(AIPlugin({ provider: 'your-ai-provider', apiKey: 'your-api-key'}));See AI Integration for provider setup and supported AI features.
Custom Asset Sources#
Connect external asset libraries like Unsplash, Getty Images, or your own content management system. Asset sources let users browse and insert content from any source.
import UnsplashAssetSource from '@imgly/plugin-unsplash';
// Add Unsplash integrationawait cesdk.addPlugin(UnsplashAssetSource({ accessKey: 'your-unsplash-access-key'}));See Custom Asset Sources for integration patterns.
Discover More Plugins#
Explore the full plugin ecosystem in the IMG.LY plugins repository. Available plugins include:
- Vectorizer – Convert raster images to vectors
- Design Presets – Pre-built design templates
- Social Media Templates – Platform-specific sizing
- And more – Check the repository for the latest additions
Key Capabilities#
The Video Captions starter kit includes everything needed for professional video captioning.
AI Caption Generation
Auto-generate accurate captions using ElevenLabs Scribe V2 speech-to-text technology.
Caption Styling
Customize fonts, colors, backgrounds, alignment, and positioning with professional presets.
SRT Import/Export
Import existing SRT subtitle files and export captions in standard subtitle formats.
Timeline Control
Precise timing controls for caption start time and duration with visual timeline feedback.
Caption Inspector
Dedicated inspector panel for editing caption text, timing, styling, and positioning.
Video Export
Export videos with embedded captions to MP4 format with customizable quality settings.
Troubleshooting#
Editor doesn’t load#
- Check the container element exists: Ensure your container element is in the DOM before calling
create() - Verify the baseURL: Assets must be accessible from the CDN or your self-hosted location
- Check console errors: Look for CORS or network errors in browser developer tools
Autocaption doesn’t work#
- Check fal.ai API key: Verify your API key is configured correctly in the proxy URL
- Verify proxy URL: Ensure
VITE_AUTOCAPTION_PROXY_URLpoints to a working proxy server - Check network requests: Look for failed requests to the fal.ai API in DevTools Network tab
- Video has no audio: The autocaption feature requires audio to transcribe
Caption Inspector doesn’t open#
- Check video captions feature: Ensure
features/videoCaptionsEnabledis set totruein settings - Verify caption mode: The inspector opens automatically in
autocaptionmode - Open manually: Use
cesdk.ui.openPanel('//ly.img.panel/inspector/caption')to open it programmatically
SRT import fails#
- Check file format: Ensure the SRT file follows standard subtitle formatting
- Verify encoding: SRT files should use UTF-8 encoding
- Check timestamps: Timestamps must be in the format
HH:MM:SS,mmm --> HH:MM:SS,mmm
Captions don’t appear in export#
- Wait for caption rendering: Ensure captions are visible in the editor before exporting
- Check caption timing: Verify caption start times and durations are correct
- Export as MP4: Captions are embedded in the video during MP4 export
Video export takes too long#
- Reduce resolution: Lower output resolution significantly improves encoding speed
- Shorten clip length: Longer videos take proportionally longer to encode
- Check system resources: Video encoding is CPU-intensive; close other applications
Watermark appears in production#
- Add your license key: Set the
licenseproperty in your configuration - Sign up for a trial: Get a free trial license at img.ly/forms/free-trial
Next Steps#
- Configuration – Complete list of initialization options
- Serve Assets – Self-host engine assets for production
- Actions – Build custom export and save workflows
- Theming – Customize colors and appearance
- Localization – Add translations and language support