Swiss Ephemeris: REST API vs pyswisseph (When to Use Which)

Almost every “how do I do astrology in code” answer points at pyswisseph — the Python bindings for the Swiss Ephemeris C library. It’s excellent. But it’s a C extension you compile, data files you ship, and chart logic you write yourself. A REST API removes all three. Here’s an honest comparison so you pick the right one.

pyswisseph exposes the raw Swiss Ephemeris functions: swe_calc_ut, swe_houses, and friends. You get maximum control and offline operation. In return you take on:

• A C extension. It compiles on install — fine on your laptop, a recurring source of CI and Docker friction (build tools, wheels per platform/arch).
• Ephemeris data files. For full JPL precision you ship the .se1 files (tens of MB). The built-in Moshier mode avoids them at a small precision cost.
• The astrology layer. swe_calc_ut returns a longitude. House systems, aspect detection, dignities, chart assembly — you build all of it.
• Licensing. The Swiss Ephemeris is dual-licensed (AGPL or a paid professional license). If you ship it inside a closed-source product, that’s a question you have to answer.

AstroWay runs the same Swiss Ephemeris (compiled to WebAssembly server-side) behind HTTP:

from astroway import Astroway, BirthData

aw = Astroway(api_key="aw_live_...")
chart = aw.chart.compute(BirthData(
    date="1990-07-14", time="14:30:00", timezone_offset=3,
    latitude=50.45, longitude=30.52,
))

No compile step, no data files, no house-system code — and the licensing of the engine is the provider’s problem, not yours. You add a network round-trip and an API key, and get 722 endpoints (houses, aspects, synastry, transits, Vedic, Human Design) instead of just planet longitudes.

• Use pyswisseph for an offline CLI, a notebook, a desktop app, or when you need the library to run with no network and you’re comfortable owning the data files, the chart code and the license question.
• Use a REST API for a web app, a serverless function, or any service where you don’t want a C build in your pipeline or ephemeris files in your image — and where you’d rather call synastry or transits than implement them.

Both are the same math. The choice is about what you want to own.

• Swiss Ephemeris accuracy benchmarks
• Natal chart API in Python
• Getting started
• Pricing