Use T1 in Your AI Assistant
T1 is also an MCP server: add it to an MCP-capable assistant (Claude, ChatGPT, Cherry Studio, and more) and ask in plain language—“how many 600×400×350 boxes fit in a 40HQ?” The count, fill rate, and a 3D loading image come back in the chat. Free, no sign-up, no API key.
https://mcp.cuberithm.com/mcp (Streamable HTTP) — what is MCP?
Setup and example questions
- In Claude (Desktop or claude.ai): Customize → Connectors → Add custom connector → paste the address above.
- In Cherry Studio: Settings → MCP Servers → Add → choose “Streamable HTTP” → paste the address → enable.
- “How many 600×400×350 cartons fit in a 40HQ container?” — presets include 20/40/45 GP·HQ, 53HC/48HC, reefers, and EU pallet-wide; the 3D view shows the exact interior dimensions used.
- “How many 600×400×350 boxes fit in a 2000×1000×1000 space?” — any rectangular space: a pallet area, leftover container space, or a larger box.
- “Load 120 cartons toward the front of a 40-ft container and show the plan.”
T1 WebAPI for Developers
For direct integration, send one carton size plus a container preset or explicit space dimensions to POST https://mcp.cuberithm.com/api/t1/pack. The response returns the count, fill rate, placed loading regions, and a shareable 3D image URL. Free, no sign-up, no API key.
Published T1 Benchmark
10,000 fixed one-carton-type 40HQ cases, updated October 16, 2026: every T1 loading plan published and checked by an independent checker, a head-to-head with a free planner, and spot checks of public loading tools. All 10,000 Deeper-search plans pass the checker, with every carton fully supported; they hold more cartons than a transparent six-orientation baseline in 7,816 cases, 2.512% more on average.
Command line
A Windows command-line copy of the same loading calculation, for checking the published benchmark numbers on your own PC.