v1 API · Python · Node · REST

Plant simulation and digital twins, as an API.

Upload a 3D layout, attach equipment models, stream live machine state and run discrete-event what-if simulations. One REST API and two SDKs, built for the engineers who wire factories and facilities together.

Read the quickstart See pricing

~/plant-twin

$ pip install steamforge
Successfully installed steamforge
$ sf login
✓ Authenticated · workspace: line-ops
$ sf twins create line3 --scene ./line3.usd
✓ twn_line3 · 42 assets · 9 stations
$ sf sim run twn_line3 --scenario shift_8h
simulating 8h0m … done in 38s (0.6 sim-min)
throughput   312 units
utilization  81.4 %
bottleneck   ST-07 Welding · queue 14

Illustrative product interface

01 · Quickstart

Run your first simulation

Set SF_API_KEY, then start a shift simulation against a twin. Same call, three ways.

curl

curl https://api.steamforge.buzz/v1/simulations \
  -H "Authorization: Bearer $SF_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "twin": "twn_line3",
    "scenario": "shift_8h",
    "duration": "8h",
    "seed": 42
  }'

run.py

from steamforge import Client

sf = Client()  # reads SF_API_KEY

run = sf.simulations.create(
    twin="twn_line3",
    scenario="shift_8h",
    duration="8h",
)
result = run.wait()
print(result.throughput, result.bottleneck)

run.mjs

import { SteamForge } from "@steamforge/sdk";

const sf = new SteamForge(); // reads SF_API_KEY

const run = await sf.simulations.create({
  twin: "twn_line3",
  scenario: "shift_8h",
  duration: "8h",
});
const result = await run.wait();
console.log(result.throughput, result.bottleneck);

02 · Reference

Endpoints and capabilities

Scenes, models, live data and simulation, in that order. Every resource is versioned and JSON in, JSON out.

How each capability works →

POST /v1/scenes Upload an OpenUSD, glTF or STEP layout and get a navigable 3D scene.
POST /v1/twins Bind a scene to assets, signals and process logic to create a twin.
POST /v1/models Register an FMI 2.0 / 3.0 FMU or reduced-order model for a machine.
POST /v1/connectors Configure an edge connector for OPC UA, MQTT or Modbus TCP sources.
PATCH /v1/twins/{id}/assets/{asset}/state Push asset state directly when you already have the data in hand.
GET /v1/twins/{id}/stream Subscribe to live state over Server-Sent Events for dashboards.
POST /v1/simulations Run a discrete-event simulation of a shift, week or custom horizon.
POST /v1/scenarios Fork a twin, change layout, cycle times or staffing, and compare runs.
GET /v1/simulations/{id}/results Throughput, utilization, queue lengths and bottleneck ranking.
POST /v1/webhooks Get notified when runs finish or live state crosses a threshold.

03 · Limits and billing

Metered on what you actually run

You pay for simulated time, telemetry messages and active twins. A sim-minute is one minute of compute time spent simulating, not one minute of simulated factory time.

Plan limits

Plan limits
Unit	Sandbox	Team	Enterprise
Price	$0	$490 / month	Contact sales
Active twins	1	10	Custom
Assets per twin	500	10,000	Custom
Sim-minutes included / month	300	5,000	Committed volume
Max single run	10 min	120 min	Custom
Concurrent runs	1	10	Custom
Telemetry messages / month	1M	50M	Committed volume
State history retention	7 days	90 days	Configurable
API rate limit	60 req/min	600 req/min	Custom
Overage	Hard stop	$0.06 / sim-min	Contract rate

Full overage rates and billing rules are on the pricing page.

04 · FAQ

Common questions

Is this a digital twin or a 3D dashboard?

Both pieces are there, but the difference matters. A 3D view or an IIoT dashboard alone is not a twin. A SteamForge twin is a scene, the equipment and process models behind it, and live operating data tied together by one information model, so you can simulate from the current state instead of a blank one.

Which file formats and protocols are supported?

Scenes: OpenUSD, glTF and STEP. Equipment models: FMI 2.0 and 3.0 FMUs. Live data: OPC UA, MQTT and Modbus TCP through the edge connector, or any source through the REST state endpoint.

Do we have to open our OT network?

No. The edge connector runs inside your network and only makes outbound connections. It is read-only by default; writing back to control systems is off unless you explicitly enable it on an Enterprise deployment.

Where is our data stored?

Team workspaces are hosted in US regions. Enterprise customers can choose a dedicated deployment or run the platform in their own environment. Your scenes, models and results remain yours.

Does SteamForge replace Plant Simulation or AnyLogic?

Not necessarily. Desktop tools like Siemens Plant Simulation and AnyLogic are strong for detailed modeling work. SteamForge is for teams that want simulation and live state behind an API, inside their own apps and pipelines. Many exported models can be brought in as FMUs.

Get a Sandbox key

Free, no credit card. Email us with your team name and the formats you work with, and we will send a key for one twin and 300 sim-minutes a month.

Request a key Talk to sales