Your creative workspace
Start creating with free tools, right in your browser.
Make something amazing today ✨
Edit video, design artwork, polish photos and work with audio in one workspace. No paid APIs or uploads.
Saved designs
Browser storageTap an element to select. Drag to move it. Change the selected element using the panel. Designs are saved in this browser.
The first version supports text and rectangle layers. More effects, templates and collaboration come later.
Picture templates
Choose a layout for your photoOne-click photo looks
Apply a filter preset, then fine-tune itChoose a picture template, customise the look, then export. All photo editing runs in your browser.
Tip: try Polaroid or Film Frame with your own caption. Templates use your imported photo, not stock images.
Adjust picture
For best quality, use the original photo. Export size varies by chosen template.
Your original file stays on your device.
Drag this playhead to jump across trimmed clips. Press Play sequence to continue from the chosen time. Original media remains unchanged.
Sequence preview plays selected ranges in order. Full sequence WebM export records in real time on supported browsers. Storyboard JSON saves order and trims, but not the original media: reselect your original files when loading. No overlapping multi-track editing yet; one background-music track is supported.
Timed titles
Subtitles (SRT)
Word highlighting divides each caption evenly across its words. It is an approximate animation, not word-aligned speech recognition.
Generate caption segments from a transcript (no AI)
Paste a transcript. CreatorOS divides its words across the sequence duration, assigning evenly spaced timings. Check and correct the timings against speech before export; this is not automatic transcription or alignment.
✨ Auto-generate captions with local AI (beta)
Speech recognition runs on your device using Whisper Tiny English through Transformers.js. On first use, the browser downloads a sizeable free model from Hugging Face (and the library from a CDN). No audio is sent to a transcription API. After assets are cached, it may work offline; not guaranteed. Best tested in desktop Chrome or Edge over localhost/HTTPS. English only.
This beta creates timestamped captions only when the model returns timestamps. Model timestamps are approximate and should be checked. The generated captions can be edited and exported to SRT. Files and audio are processed locally by the browser; external model files are downloaded on first use.
To export: first import a video in Video Studio, choose WebM (recommended), click Export full sequence at the top, and keep this tab open until a file downloads. If it fails, an error appears here and at the bottom of the screen.
Audio edits and WAV rendering run entirely on your device.