Your AI sees what you see

One speech.
Everything you say and copied.

One prompt for any AI. Saved together. In order.
Revisit, search, share. File-system only.
1 · You talk + copy
10:41:02
Spoke · 18s
So the issue happens when...
10:41:40
Copied · article excerpt
Chrome
10:42:05
Copied · Claude Code session
Claude Code · a3f9c2
10:42:19
Spoke · 7s
Now compare it with the docs...
10:42:48
Copied · 7 images
Figma · hero-v3 frames
10:43:08
Annotated · screenshot
edited here · arrows + counters
10:43:24
From iPhone · 3 photos
scanned the QR
10:43:41
GIF
Copied · product demo
Chrome · demo-loop.gif
10:43:55
Copied · 10 files
Finder · ~/specs
10:44:25
Spoke · 9s
Let's try a simpler approach...
saved togetherin order
optionOS
One prompt
Everything from one session.
speech + copies, merged chronologically
3 · Prepared input
Ready to paste anywhere
In order · labeled for your agent
Spoke → transcript18s
article excerptChrome
Claude Code sessiona3f9c2
Spoke → transcript7s
7 images → bundleFigma
annotated shotedited
3 photosiPhone
GIF GIF → keyframes9 frames
10 files → pathsFinder
Spoke → transcript9s
All set. Paste it into any AI or tool.
ChatGPT Claude Gemini Perplexity Cursor Copilot Grok Any AI
Watch it work

Ten minutes. Everything, end to end.

Recorded with optionOS itself — the recording pipeline you see is the product.
Full walkthrough Turkish · 10 min
Don't feel like watching? Copy the transcript — search it, or paste it into any AI and ask in your language.
optionOS optionOS
Explore more

Everything you need while you work.

Talk, capture, search, revisit, and paste — all saved together, in order.
1Compose⌥ A
optionOS HUD — recording pipeline and copied items
  • speak — your thoughts line up in order
  • start / stop anywhere
  • copies merge into your session
more
  • works in a crowd — keeps your voice, drops the rest
  • two people dictating side-by-side don't mix
  • recordings searchable & replayable
  • 6 engines · 16 languages · runs fully local
2Capture
Everything you copy, saved with source and time
  • every copy saved — code, links, shots, PDFs
  • source app + time on every item
  • dictations land here too
more
  • screenshot ⌥⇧S · GIF ⌥⇧G · scrolling capture ⌥⇧X
  • GIFs carry keyframes — AI agents read them
  • auto copy mode · undo / redo your grabs
  • images appear inside your transcript
3Annotate
Markup editor — arrows, counters and text on any capture
  • stars survive delete
  • tags & keywords
  • OCR + instant preview
more
  • 12 tools — arrows, counters, text, blur, spotlight
  • save · copy · share from the editor
4Search & revisit
Search your context — live filter chips, dictation transcript and session links
  • /type @source #keyword — search with live filters
  • jump back to the exact Claude Code session
  • star · tag — instant collections
more
  • timestamps link copies ⇄ speech ⇄ sessions — your own graph
  • tell your agent — it finds that exact conversation
  • plain · badge · detail views per item
  • paste inline or hand to a subagent
5From your phone
Scan the QR with your phone — photos land on your Mac
  • scan the QR — it lands on your Mac
  • photos, notes, links
  • no AirDrop dance
6Global Inspector⌥ C
Global Inspector — hover any element in any app, copy its identity and HTML with one key
  • hover anything — your AI sees exactly what you're focused on
  • ⌥C mid-sentence — the element's full HTML drops into your session
  • works Mac-wide — Chrome, Figma, Finder, every app
more
  • name, role and CSS selector appear live as you point
  • no DevTools, no inspect-element detour — your flow never breaks
  • web elements arrive as real HTML, native ones with their exact identity
© optionOS · Your AI sees what you see. optionOS What's New hello@optionos.app