About SayScreen
One sentence in, a narrated tutorial video out — and a video you can fix by changing a line instead of recording it again.
Why it exists
A short video is the fastest way to show someone how software works. Making one is slow: write a script, clear the desktop, record take after take, then zoom, cut and caption by hand. And the moment the interface changes, the video is out of date and the work starts over.
SayScreen does the recording and the editing. You point it at the page where the lesson starts and describe what to teach. It drives a real browser through your app, writes and voices the narration, and edits the video with smooth zooms, a gliding cursor and captions. When your app changes, you run the same prompt again.
How it works
- Claude Opus 5.5 from Anthropic directs: it reads each screen, decides where to click and writes a line of narration for every step.
- ElevenLabs voices the script — a stock voice or your own cloned voice.
- SayScreen records in Chrome at 2× scale and edits the footage the way a careful human would, then hands you an MP4 and an .srt caption file.
Who builds it
SayScreen is built by WesGPT — the YouTube channel where Wes explores AI and automating everything. Ideas, questions and bug reports go straight to the person building it.
Your next tutorial is one sentence away.
Point SayScreen at your app, say what to teach, and get a narrated, edited video back. Free to start.
Free plan · No credit card · Cancel anytime