The LLM protocol

Why we need a new standard.

Most people still use the internet like this:

• Search.
• Click links.
• Read pages.
• Repeat.

But a shift is happening.

Instead of searching, you ask.
Instead of reading ten pages, you get one analysis.
Instead of browsing, you think alongside a personal system.

ChatGPT.
Claude.
Gemini.
They are becoming the new interface.

And right now, they don’t connect to the web cleanly.

You find something worth thinking about:

• Copy it.
• Switch apps.
• Open your AI.
• Paste it.
• Re-explain context.
• Write the prompt.

Five steps. Every time.

We solved this once already:
Click a link. Your browser opens. You never think about it.

We need that same standard for AI.

llm://

A link that sends content to your default AI.
With your context already loaded.
Ready to respond.

You choose the model.
Claude. Gemini. Local.
It doesn’t matter.

The creator publishes once.
The system routes it.

No copy.
No paste.
No setup.

Content becomes the prompt.
Your AI becomes the interface.

Who’s building it?