AI with Eyes
AI has read everything and seen almost nothing. OpenEyes is the live sensor layer for machine intelligence — every camera on earth, in one directory, callable by any agent.
Make every piece of the earth visible to AI.
The world is full of views no agent can reach — a snowed-in pass, a reef coming alive at dawn, a line snaking out the door. If the view doesn't exist yet, we'll pay to bring it online: point a camera at something worth seeing and agents rent your eyes by the frame. The map fills itself.
Every view earns.
Add an eye. We list it, agents find it, and you get paid every time one looks through it. You keep 90%. Or keep it free if you'd rather just share.
- ▸Still image or live video
- ▸Free by default — turn on pricing any time
- ▸No Stripe until you want to cash out
Questions an agent can't answer without seeing.
- “Is Snoqualmie Pass snowed shut right now?
- “How long is the lift line at Breck?
- “Are the bears out at the salmon run?
- “Is Rincon actually breaking, or am I gonna waste a drive?
Point your agent at the planet.
Pick your client, paste the config, ask it something that needs eyes.
Claude Desktop
Anthropic · macOS + Windows- 01Open the config file
On macOS:
~/Library/Application Support/Claude/claude_desktop_config.json
On Windows:
%APPDATA%\Claude\claude_desktop_config.json
Create the file if it doesn't exist.
- 02Add the videoai server
{ "mcpServers": { "videoai": { "url": "https://api.openeye.cam/mcp" } } } - 03Restart Claude Desktop
Fully quit and reopen Claude. You'll see
videoaiin the tools menu at the bottom of the chat input. - 04Try it
Ask it anything that needs eyes:
- “What’s I-70 looking like at the Eisenhower right now?”
- “Any bears at Brooks Falls today?”
- “Is Rincon breaking?”