Know what any part should cost — and why.
ARCNM extracts every feature and tolerance from your 3D model and drawing, simulates the manufacturing process and its setup and cycle times, then decomposes the cost across the lot-size curve — so every number on the quote traces back to the geometry it came from, not a black-box guess. Physics computes the cost structure; machine learning is applied only where it earns its place — reading drawings with vision-language models and calibrating to your shop. API-first and MCP-native, for your code, your AI agents, and your automations.
What ARCNM does
ARCNM is physics-grounded, AI-backed cost intelligence for manufactured parts. Given 3D geometry (STEP/STP/STL), a 2D drawing (PDF/DXF), and a batch size, it returns an auditable should-cost — material, setup, cycle time, labour, overhead and unit price across a lot-size curve — and the full extracted part data behind it. Physics computes the cost structure; machine learning is applied only where it genuinely wins: reading drawings and calibrating to your shop.
Deep feature & PMI extraction
A B-rep geometry engine recognises holes, pockets, faces, bosses, bends, slots, chamfers and threads from 3D, while vision-language models read dimensions, tolerances, GD&T and surface finish from the 2D drawing (ISO 286, 14405, 1101, 5459, 7200, 16792, 21920). Every feature and all PMI come back as structured, typed, versioned data — a first-class output you can reuse for master-data management, ERP/PLM backfill or similarity search.
Manufacturing-process physics
Discipline classification (milling, turning, sheet metal, casting, wire EDM, grinding) routes the part to physics handlers that compute setup and cycle time from first principles, then a lot-size curve surfaces the unit price and the break-points at every quantity — calibrated to your shop, with a confidence interval on every number.
Built for code and AI agents
A single REST call (POST /v1/extract) plus an MCP server let your services, AI agents, and iPaaS automations price parts natively. Every quote is fully auditable back to the parameter that produced each number.
Pay only for results
€0.89 per successful extraction, €0 on failed or duplicate calls. Free to start with €15 credit and 50 calls a month — no credit card required.