01 / AI FOR PHOTONICS

Less waiting.
More discovery.

Predict photonic circuits with AI.
98% match to 3D FDTD.
Explore ideas at the speed you think.

Get your API key ↗ Read the docs

Free during alpha. No credit card.

Predicted electric field intensity in a 1550 nm silicon Y-branch splitter — FIELD PREVIEW / |E|²

1.3 s

bamsolve, full 15 µm Y-branch, fields + S-parameters

~48 min

3D FDTD (Meep, 30 px/µm, 8 cores)

~2,200×

faster

From geometry to insight.

01

Define

Bring your GDS and a layer stack. No meshing, no setup files.

02

Predict

Full complex fields, S-parameters and n_eff in ~1.3 s.

03

Explore

Cut anywhere. Watch one mode split into two, then export Touchstone for your compact models.

Field intensity across the Y-branch: one lobe at the input, two half-power lobes at the outputs

02 / API

Three calls. That's it.

Define a stack, load your layout, predict. Every result exports straight to .sNp.

Full GDS & API guide →

# pip install bamsolve
from bamsolve import Layer, Stack, bamsolve, pred

bamsolve(api_key="bsk_...")

# 1. Cross-section stack
stack = Stack(clad=1.444, layers={
    (1, 0): Layer(n=3.48, z=(0.0, 0.22),
                  angle=5.0),  # sidewall angle (°)
})

# 2. Apply it to a GDS
dev = stack("ebeam_y_1550.gds")

# 3. Predict fields + S-parameters
res = pred(dev, inputs={((1, 10), "TE", 1.55)})
res.export_s("ebeam_y_1550.s3p")

Get your API key.

Free for everyone during the public alpha.