Live today: 8 processes — 3D printing, CNC, sheet metal, laser cutting, stickers, paper print, apparel · US shipping

The agent-native procurement network
for custom fabricated parts.

Your agent designs a part — a bracket, an enclosure, a panel, or the merch to go with it. AFN returns ranked offers — price, delivery date, and the vendor's product preview — then checks out with one call. No human UI. No hidden fees. Built exclusively for AI agents.

Read the docs→ I'm an agent→
agent session — api.agenticfabricationnetwork.ai
# any agent, four calls, zero setup meetingsPOST /v1/agents                → api key, self-servePOST /v1/files                 → register artwork or STLPOST /v1/quotes                → ranked offers: price · ETA · preview
     { "process": "decal", "constraints": { "deadline": "2026-07-13",
       "submersible": true, "ship_to_zip": "78701" } }POST /v1/checkout              → hosted payment link for the user# after payment, AFN places the vendor order automatically
8fabrication processes live on the network today
secondsfrom request to ranked, purchasable offers
$0hidden fees — the price shown is the price paid

Why AFN

One protocol. Every process.
Zero setup meetings.

ONE FUNNEL

Every process, one protocol

Eight processes today; PCBs and injection molding are a process field away. Agents never learn a new tool.

CONSTRAINTS

"Must survive pool water"

Deadline, outdoor-rated, submersible, budget — first-class request fields that narrow offers instantly.

PREVIEWS

See it before you buy it

Every offer carries the vendor's own product render of your artwork with those exact options.

CHECKOUT

All-in pricing, one call

The price shown is the price paid. AFN is the merchant of record and places the vendor order after payment.

For agents & agent builders

AFN is machine-first: capability discovery, an OpenAPI spec, an llms.txt, and an MCP server (works in ChatGPT as an app, and in Claude, Cursor, or any MCP client). AFN speaks UFP — the open Universal Fabrication Protocol. Point your agent at any of these:

/llms.txt /openapi.json /v1/capabilities /docs/protocol.md