# Climate Simulation API: Python and TypeScript SDK · infrared.city

> Run wind, wind comfort, UTCI, solar and daylight simulation from Python or TypeScript. SDK, REST API and agent skills. Get a key and run in minutes.

Source: https://infrared.city/products-api

Run climate simulations from your own code. Wind, heat, sun and daylight for one site or a whole city.

[Docs](https://infrared.city/docs/sdk/) [Playground](https://platform.infrared.city/sdk/playground) [Get your API key](https://platform.infrared.city/)

![The API playground: wind speed inputs next to a 3D wind result on a demo site](https://infrared.city/assets/media/crop-api-playground-wind.webp)

Wind speed on a demo site, from the API playground.

**Thousands** of simulations per second

**44 s** 400 km² of Berlin, end to end 400 km²: perceived heat (UTCI), one summer day at 1 m, four 100 km² runs at once, 23 Oct 2026.

**12 models** Wind, heat, sun, daylight, energy

**3.20M+** simulations computed, as of 15 Nov 2026 Simulations computed include our own studies and stress tests.

### Start building

Native SDKs for **Python** and **TypeScript**.

- 900 free tokens
- No credit card

Install the SDK, set your key and run a first analysis in a few lines. The same code works for one block and for a whole city.

- **Tiling built in:** the SDK cuts large areas into tiles and merges the result. How it works

- **Geometry in one call:** `client.buildings.get_area(area)` fetches the buildings for an area.

- **Run and export:** JSON grids, PNG, GeoTIFF, 3D Tiles.

- **Recipes:** web app, Grasshopper, notebooks and a safe key proxy. See the recipes

[Full guide and reference in the docs](https://infrared.city/docs/sdk/)

Python

```
pip install infrared-sdk
```

TypeScript

```
npm install @infrared-city/infrared-sdk
```

Python
TypeScript

```
# pip install infrared-sdk   (set INFRARED_API_KEY)
# Buildings outside our served cities: pip install "infrared-sdk[geodata]"
from infrared_sdk import InfraredClient, SvfModelRequest
from infrared_sdk.analyses.types import AnalysesName

area = {"type": "Polygon", "coordinates": [[          # a block in Vienna
    [16.369, 48.207], [16.373, 48.207], [16.373, 48.209],
    [16.369, 48.209], [16.369, 48.207]]]}

with InfraredClient() as client:
    buildings = client.buildings.get_area(area)
    result = client.run_area_and_wait(
        SvfModelRequest(analysis_type=AnalysesName.sky_view_factors),
        area, buildings=buildings)
grid = result.physical_grid()   # sky view factor, 0 to 100 %
```

```
// npm install @infrared-city/infrared-sdk   (set INFRARED_API_KEY)
import { InfraredClient, initializeCore, areaGridValuesF32, type AreaResult } from "@infrared-city/infrared-sdk";

await initializeCore();
const area = { type: "Polygon", coordinates: [[   // a block in Vienna
  [16.369, 48.207], [16.373, 48.207], [16.373, 48.209],
  [16.369, 48.209], [16.369, 48.207]]] };

const client = new InfraredClient({ apiKey: process.env.INFRARED_API_KEY });
const buildings = await client.buildings.getBuildingsInArea(area);
const result = await client.runAreaAndWait(
  { analysisType: "sky-view-factors" }, area, { buildings });
const grid = areaGridValuesF32(result as AreaResult);   // 0 to 100 %
```

The SDK

### Very large areas, one function call

The SDK cuts your area into 512 m tiles, runs each tile as its own job
and merges the results into one result.

- **Any size.** A block, a district or a whole city.

- **In parallel.** 100 km² takes 37 s, not 100 times as long as 1 km².

- **Reliable.** A failed tile is sent again. No holes, no double bill.

Standard outputs

- JSON grids
- PNG maps
- Facade values
- OBJ and dotbim
- GeoTIFF
- 3D Tiles

A sensor reads one corner, infrared.city simulates every street

Sensors show it is hot, now, at a few points. infrared.city is the simulation side: it shows where, on every street, tomorrow, in 2050 and after you plant the trees.

- Sensors Simulation (infrared.city)

- Measure → Predict

- A few points → Every street

- Now → The next 72 hours, and scenarios

- Nothing to test → Test adaptation and scenarios

### For enterprise and platforms

Built for teams that run it inside their product.

One of the largest AEC software platforms runs our engine, since 2026. More than 2,500 accounts use it.

![The live climate layer over Hong Kong](https://infrared.city/assets/stills/live-layer-hongkong.webp)

Example: Hong Kong. Buildings, trees and terrain © Lands Department, HKSAR.

- **Reliable at scale** Failed jobs rerun automatically.

- **EU hosting and GDPR** Frankfurt, data deleted after 24 h.

- **Your terms** DPA, OEM licence, SLA.

- **On-prem possible** Capacity raised with you.

How does this compare with ENVI-met, SOLWEIG or PALM-4U? See [when to use which tool](https://infrared.city/products-compare).

The agent skill

### Let your AI agent write the code

The skill teaches your coding agent the SDK and its recipes: web apps,
3D facade viewers, Grasshopper, notebooks and secure backends. Your agent
runs real simulations with your key and does not invent numbers.

In Claude Code or Cursor

```
/plugin marketplace add Infrared-city/infrared-skills
/plugin install infrared@infrared-skills
```

Codex, Copilot, Windsurf: read [AGENTS.md](https://github.com/Infrared-city/infrared-skills/blob/main/AGENTS.md).

[Plugin marketplace on GitHub](https://github.com/Infrared-city/infrared-skills)

You ask

Build me a small web app: I draw an area, press Run, and see thermal comfort on a map.

### Build an app in less than an hour

Pick a recipe. Each one is a working example in our public repository;
your agent reads it and adapts it to your site, your data and your stack.

- [**Web app or client portal** Draw an area, press Run, see the result on a map.](https://github.com/Infrared-city/infrared-skills/tree/main/cookbook/apps/map-grid)
- [**3D facade viewer** Sun on every wall and roof, the value on hover.](https://github.com/Infrared-city/infrared-skills/tree/main/cookbook/apps/facades-3d)
- [**Grasshopper script** Wind, sun or comfort in Rhino 8, baked into your model.](https://github.com/Infrared-city/infrared-skills/blob/main/plugins/infrared/skills/use-infrared/references/recipes/grasshopper.md)
- [**Notebooks and batch studies** Ten ready notebooks for variants, heat, wind and cost.](https://github.com/Infrared-city/infrared-skills/tree/main/cookbook/notebooks)
- [**Many users, key kept safe** A Worker or a FastAPI backend keeps your key on the server.](https://github.com/Infrared-city/infrared-skills/tree/main/cookbook/apps/cloudflare-proxy)
- [**Digital twin with 3D Tiles** Your results as 3D Tiles, ready for Cesium.](https://infrared.city/docs/sdk/)

### Built with infrared.city

Our own apps first, then our demos, then what teams built in five days at
the [infrared.city Buildathon](https://hackathon.infrared.city/projects)
in May 2026. Point at or tap a card for more.

-
![The infrared.city platform: screenshot](https://infrared.city/assets/showcase/platform.webp)
Our app
**The infrared.city platform**
by infrared.city

-
![Hong Kong climate twin: screenshot](https://infrared.city/assets/showcase/hong-kong-twin.webp)
Our app
**Hong Kong climate twin**
by infrared.city

-
![Design & compare: screenshot](https://infrared.city/assets/showcase/design-explorer.webp)
Our demo
**Design & compare**
by infrared.city

-
![Construction heat planning: screenshot](https://infrared.city/assets/showcase/construction-heat.webp)
Our demo
**Construction heat planning**
by infrared.city

-
![Streets & public spaces: screenshot](https://infrared.city/assets/showcase/climate-corridor.webp)
Our demo
**Streets & public spaces**
by infrared.city

-
![Marathon & event routes: screenshot](https://infrared.city/assets/showcase/route-analyzer.webp)
Our demo
**Marathon & event routes**
by infrared.city

-
![Portfolio performance: screenshot](https://infrared.city/assets/showcase/portfolio.webp)
Our demo
**Portfolio performance**
by infrared.city

-
![Treeat: screenshot](https://infrared.city/assets/showcase/treeat.webp)
Buildathon 2026
**Treeat**
by Mari

-
![CoolSpend: screenshot](https://infrared.city/assets/showcase/coolspend.webp)
Buildathon 2026
**CoolSpend**
by Rafik El Khoury

-
![Renovation Map Toolkit: screenshot](https://infrared.city/assets/showcase/renovation-map.webp)
Buildathon 2026
**Renovation Map Toolkit**
by Medinaura

-
![Detour: screenshot](https://infrared.city/assets/showcase/detour.webp)
Buildathon 2026
**Detour**
by xBuild

-
![HeatGuard: screenshot](https://infrared.city/assets/showcase/heatguard.webp)
Buildathon 2026
**HeatGuard**
by HeatGuard

-
![WindSite: screenshot](https://infrared.city/assets/showcase/windsite.webp)
Buildathon 2026
**WindSite**
by Team Turbine

-
![Cairo PEDCOMF: screenshot](https://infrared.city/assets/showcase/cairo-pedcomf.webp)
Buildathon 2026
**Cairo PEDCOMF**
by Antony Hozeen

-
![simSPACE and Tree SimCity: screenshot](https://infrared.city/assets/showcase/tree-simcity.webp)
Buildathon 2026
**simSPACE and Tree SimCity**
by CCAE / Cork Centre for Architectural Education

-
![CoolTure: screenshot](https://infrared.city/assets/showcase/coolture.webp)
Buildathon 2026
**CoolTure**
by CoolTure

-
![CIRCE: screenshot](https://infrared.city/assets/showcase/circe.webp)
Buildathon 2026
**CIRCE**
by Team NS

-
![CoolRun AI: screenshot](https://infrared.city/assets/showcase/coolrun-ai.webp)
Buildathon 2026
**CoolRun AI**
by Yusuf

-
![UrbanShade: screenshot](https://infrared.city/assets/showcase/urbanshade.webp)
Buildathon 2026
**UrbanShade**
by Rashi Desadla
