Steltic

Watch Steltic work

What is Steltic?

Steltic is a family of free, open-source structural design assistants for engineers: an agentic harnessi that drives an AI LLM through the full steel design workflow — one assistant for hot-rolled steel (AISC 360/341) and one for cold-formed steel (AISI S100/S240/S400). Both live on GitHub; small capped Demos run here on our servers.

Steltic has no LLM of its own — it is an LLM harness that guides your LLM through the structural design workflow but generates no engineering content of its own; every model, calculation, figure, and report is produced by your LLM. You simply input your API URL and API key at the beginning of your session (you pay the LLM costs).

The harness is designed to lead the LLM through a first pass, then enters a traditional "chat" format, wherein the building is now in context and the user can ask the LLM to make any changes.

Quick Start — 61 worked example buildings ship across the two assistants (36 AISC + 25 CFS). Run one free in a Demo, or install locally and design your own.

What LLM should I use?

Steltic has been internally tested thoroughly with Claude Fable 5 and ChatGPT 5.6 Sol. These frontier models perform exceptionally well and typically complete the full design with no human-in-the-loop. Notably, ChatGPT 5.6 Sol systematically designs quite conservatively in the initial no human-in-the-loop run (this simply requires the user to prompt an optimization run following the first completion). Limited testing has shown Opus 4.8 to also be very good. Without doubt, Fable 5 is the best choice for excellent results.

Open source — Zai GLM 5.2 and DeepSeek 4 Pro have been extensively tested; however they perform substantially less well, sometimes forget the agent instructions, sometimes "spin their wheels" on unimportant issues, and generally require substantial human-in-the-loop support.

Other LLMs — any LLM for which a user has an API key can be used in the harness; however, as with all AI products, less intelligent models will give less intelligent results. Nonetheless, you can test any model you wish.

You can't give us your data — so we can't have it

The free Demos run our own published example buildings only. The brief is fixed, file uploads are disabled, and the server refuses anything that isn't one of our examples.

The one thing of yours a Demo touches is your AI provider API key. It is held in server memory for the length of your session, never written to disk and never logged, and discarded when the session ends. You pay your provider directly; we never see your billing. Data privacy for your LLM usage is between you and your provider.

Your run's working files sit in an isolated workspace on an ephemeral cloud serveri (Google Cloud Run) that is cleared automatically when the instance recycles — so download your package before you go.

We record usage metadata only: an anonymous one-way hash of your IP (to enforce the Demo limits — we never store the address itself), run times, durations, outcomes and token counts. No accounts, no name, no email, no password. See the Privacy Policy.

Don't LLMs hallucinate?

Your LLM is fully grounded to the latest specifications and design procedures. The AISC assistant has access to RAG vector databases of AISC 360, 341 and 358 (2022), an examples database of OpenSees models, two databases of the OpenSees documentation, and the AISC Steel Construction Manual Design Examples of 168 design examples (V16.0). The CFS assistant is grounded the same way to AISI S100-16 (R2020) w/S2,S3, S240-20 and S400-20. These are hosted on our company's private servers, and Steltic directs your LLM to query them to ground its calculations.

What goes in, what comes out

In the Demos — input is a choice, not a document. You pick one of the example buildings from a list. The brief is locked, uploads are disabled, and the design runs once: there is no follow-up conversation and no editing the building. If your model needs more than an hour, Steltic continues automatically in the background, up to a 2-hour limit.

Running it yourself — input is anything you can describe. In the open-source versions you write the brief in text (stories, bays and spacings, loads, seismic, system), attach drawings or PDFs if your model reads images, and keep working with the agent afterwards. After the initial run your building is in context, so you can change the geometry, change the seismic LFRS, optimise the member sizes based on the drift, etc. Simply chat with your LLM to investigate any alternative designs.

Outputs are the same either way. A structural engineering report (report.html) written to the level a senior engineer needs for review, the analysis model files, an interactive 3D viewer of that model, the conversation history and the reasoning trace where your model exposes one. The 3D viewer is how you verify the modelling. In the open source versions you can modify the report parameters to include anything you think is useful for reviewing purposes. You get consistent data output for every building, every design, every time.

Download before you leave. Demo sessions run on ephemeral storage and are cleared automatically — once the session ends, the files are gone.

Browse the worked examples ↑

Steltic is open source

Both assistants are MIT-licensed: github.com/Steltic/steltic (AISC 360/341) and github.com/Steltic/steltic_cfs (AISI S100/S240/S400). Install and run them on your own machine: your own buildings, no time limits, no third party but the LLM provider you choose.

RAG. For copyright reasons the vector databases of the AISC and AISI standards are not open-sourced. Two options:

1. No RAG server (default — not recommended). The agent relies on its own cited standards knowledge. Good models do respectably; grounded runs are better.

2. BYO RAG server. Run your own and build the standards collections from your own licensed copies — the chunking and ingestion starter kit is in each repo. The freely-redistributable collections are downloadable from the Releases pages.