STL Emboss

Personalised gifts with Claude in Chrome

Ask an agent to make the gift and write on it. Claude in Chrome, a browser extension from Anthropic, drives STL Emboss the way a keyboard user does, in your own browser.

Open STL Emboss

A 16-second sped-up film: Claude in Chrome makes a 20 cm vase, opens it in STL Emboss, picks a script font, adds a heart and a line of words, and embosses them onto the model.

What to say

Name the thing, the words and the place. For example:

Open stlemboss.com, open this STL, add "Happy birthday" on the front in a script font, sunk 0.5 mm, and download it.

The agent types, places, cuts and saves while you watch, and you can take over at any moment. No model at hand? Ask it to make one first, as it does with the vase in the film. It works in your own tab: the page has no server to send anything to, and your model is never uploaded.

Check the result in your slicer before you print. An agent can be wrong about a place on the model, and the page is offered as is.

Why an agent can do this

Every control on the page has a name, and every result is written out as a full sentence with its numbers: the same things a screen reader uses. So an agent reads what happened instead of guessing from a picture. To repeat the job without clicking, it can write recipes.

For developers

There is no server API. llms.txt tells an agent how to use the page. Where a browser offers WebMCP (document.modelContext, behind a flag in Chrome today), the page also registers five tools: describe_model, add_text, adjust_mark, cut_mark and download_stl. The visitor still opens the model with “Choose a file”, and no tool hands the model to the agent.

Try it on your own modelFree, no account, and your file never leaves your device.

Open STL Emboss