fitcxgp

Fast, all-in-one GPX/TCX/FIT toolkit.

Load, inspect, and convert GPX, TCX, and FIT files with one Python, Rust, or CLI API. Every supported input can be written to every supported output.

Why fitcxgp?

• All in one — GPX, TCX, and FIT read/write support.
• Fast — Rust-powered streaming parsers; GPX parsing is up to 31x faster than gpxpy on the real-data benchmark.
• Useful model — points, segments, telemetry, device info, moving stats, elevation gain, bounds, and track simplification.
• Reliable — FIT CRC validation, property tests, and 8,280 conversion chains verified on 920 real GPX files.
• Easy to embed — Python 3.12+ ABI3 wheel, Rust library, and CLI.

Python

uv add fitcxgp

Load once, write any format:

import fitcxgp

activity_file = fitcxgp.load("run.gpx")  # GPX, TCX, or FIT

gpx = activity_file.to_gpx()
tcx = activity_file.to_tcx()
fit = activity_file.to_fit()

Inspect an activity:

activity = activity_file.activities[0]

print(activity.point_count)
print(activity.distance_meters)
print(activity.duration_seconds)
print(activity.average_heart_rate_bpm)
print(activity.elevation_gain.uphill_meters)

Convert files directly:

fitcxgp.convert("run.gpx", "run.fit")
fitcxgp.convert("run.fit", "run.tcx")

Add device information when creating FIT output:

device = fitcxgp.DeviceInfo(
    manufacturer_id=1,
    product_id=3415,
    serial_number=1234567890,
)
fit = activity_file.with_device_info(device).to_fit()

CLI

cargo run --release -- convert run.gpx run.fit
cargo run --release -- info run.fit
cargo run --release -- detect run.tcx

Rust

use fitcxgp_core::{convert, load};

let activity_file = load("run.fit").unwrap();
println!("{} points", activity_file.point_count());
convert("run.fit", "run.gpx").unwrap();

Development

uv sync --python 3.12
uv run pytest
cargo test --workspace

Conversions preserve the shared activity model, not format-specific data such as FIT developer fields or GPX routes and waypoints.

See testing, benchmarks, and the real-data report for details.