For platforms & design tools

Climate feedback for
every massing variant.

neoBim generates the variants. Infrared scores them — wind, sun, thermal comfort, daylight — in seconds, at neighbourhood scale.

Python SDK Live on PyPI
pip install infrared-sdk

8 analyses · auto-tiling for large areas · TypeScript next

Agent skills + cookbook Open on GitHub
github.com/Infrared-city/infrared-skills

10 notebooks · skills for Claude, Cursor, Codex, Copilot, Windsurf

Talk to us Try the playground
Animated 3D urban model with floating dashboards showing wind, daylight and report panels

What you can run

Eight analyses. One call per variant.

Score every variant your users generate. Weather and context handled, results in under 10 seconds per tile — fast enough for every iteration.

Wind speed
ML CFD-style flow at pedestrian level
“Will the courtyard be too gusty for outdoor seating?”
Pedestrian wind comfort
Lawson, NEN, VDI, Davenport class maps
“Where can people linger versus walk through?”
Solar radiation
Cumulative kWh/m² at street level
“How much summer sun load hits this plaza?”
Daylight availability
Hours of useable daylight per cell
“Will the plaza feel bright in November?”
Direct sun hours
Geometry-driven sun exposure
“How long will the playground stay sunlit in March?”
Sky view factor
Percent of hemisphere visible (0–100)
“Is this canyon too enclosed to cool down at night?”
Thermal comfort (UTCI)
Felt temperature in °C
“Will pedestrians overheat on a July afternoon?”
Thermal comfort statistics
Hours per heat / cold stress band
“How many days a year is this plaza uncomfortable?”
Multi-tile microclimate simulation covering a large urban area (Tokyo)

City-scale by default

Draw the whole neighbourhood.

The SDK tiles any polygon, runs in parallel, and blends seams into one seamless map — distant buildings still cast shadows across boundaries.

Live, in-browser

Try it on the page.

Built with the SDK. In one hour.

For your dev team & AI assistants

SDK skills for
Claude, Cursor, Codex, Copilot, Windsurf.

The infrared-skills repo: one agent skill, 10 runnable notebooks, six preset cities.

  • ›SKILL.md + references
  • ›Per-analysis recipes
  • ›Async + webhooks demo
  • ›Six preset cities to remix
Browse on GitHub Read the SDK docs
Infrared agent skills orbit — teal palette, Claude / OpenAI / Cursor logos surrounded by simulation glyphs (UTCI, wind, solar, IFC, code)
Claude Code · /plugin
$ /plugin marketplace add Infrared-city/infrared-skills
$ /plugin install infrared@infrared-skills
Cookbook · clone & run
$ git clone https://github.com/Infrared-city/infrared-skills
$ jupyter lab cookbook/notebooks/

Why this fits neoBim

Your variants, scored on wind, sun, thermal comfort, and daylight.

No CFD setup, no weather plumbing, no simulation expertise on your side. One API, your users keep iterating.

Insight reports

Coming soon

From heatmaps to
conclusions.

Plain-language conclusions, ready for a client presentation. In preview now, partners later this year.

  • ›Heat-island mitigation
  • ›Massing impact
  • ›Natural ventilation
  • ›+ 7 more
Get early access

Interior models · IFC-native

Coming soon

Climate analysis,
indoors too.

Drop in an IFC, get a 3D daylight heatmap. Same SDK, same workflow.

Daylight Factor — in preview Energy — roadmap Embodied carbon — roadmap
Get early access

Pluvial flood · surface water

Coming soon

Where does the rain
go on this layout?

GPU-solved pluvial flood per massing option — depth, velocity, hazard class per cell, for today’s extreme rainfall and 2100 scenarios. Validated against municipal flood maps.

  • ›Depth & velocity per cell
  • ›10 / 30 / 100-year storms
  • ›Today vs. 2100 climate
  • ›Hazard classes (DWA-M 119)
Get early access

Get in touch

Embed Infrared
in your product.

Climate scoring for the variants neoBim already generates — let’s talk.

Talk to us Read the docs