Migrate from Hevy

Export your training history from Hevy and convert it to openweight format in under a minute.

Step 1: Export from Hevy

1. Open the Hevy app
2. Go to Profile → Settings
3. Tap Export & Import Data
4. Download the CSV file (e.g., hevy.csv)

TIP

Hevy exports everything — workouts, exercises, sets, supersets, and notes. The converter handles all of it.

Step 2: Convert

npx @openweight/cli convert hevy.csv -o workouts.json --pretty

No extra flags needed. Hevy's CSV includes weight in both kg and lbs columns, so units are auto-detected.

Step 3: Verify (optional)

npx @openweight/cli validate workouts.json

Or see full conversion stats:

npx @openweight/cli convert hevy.csv -o workouts.json --pretty --report

Example output:

--- Conversion Report ---
Source:     hevy
Rows:       3104/3104 converted, 0 skipped
Workouts:   183
Exercises:  27

What Gets Converted

Hevy field	openweight field	Notes
title	name	Workout name
start_time / end_time	date + durationSeconds	Duration calculated from timestamps
description	notes	Workout notes
exercise_title	exercise.name	Normalized when possible
superset_id	supersetId	Superset groupings preserved
reps	reps
weight_kg / weight_lbs	weight + unit	Prefers kg; auto-detects which column has data
duration_seconds	durationSeconds (set)	For timed exercises
distance_km / distance_miles	distance + distanceUnit	Prefers km
rpe	rpe	When tracked
set_type	Set type mapping	See table below

Unit Auto-Detection

Hevy exports include both weight_kg and weight_lbs columns. The converter:

• Uses weight_kg when available (stores as kg)
• Falls back to weight_lbs (stores as lb)
• Same logic for distance: prefers distance_km over distance_miles

You never need to specify units manually.

Exercise Name Normalization

Common Hevy exercise names are mapped to canonical names:

Hevy name	openweight name
Bench Press (Barbell)	Bench Press
Back Squat (Barbell)	Squat
Deadlift (Barbell)	Deadlift
Lateral Raise (Dumbbell)	Lateral Raise

Set Types

Hevy set_type	openweight set type
normal / 1	(default working set)
warmup / 2	warmup
dropset / 3	dropset
failure / 4	toFailure: true

Supersets

Hevy's superset_id column is preserved as supersetId on the exercise. Exercises with the same superset ID are grouped together, matching how they appear in Hevy.

Full CLI Reference

npx @openweight/cli convert <file> [options]

Options:
  -f, --format <format>       Source format (auto-detected from headers)
  -u, --weight-unit <unit>    Weight unit: kg or lb (not needed for Hevy)
  -o, --output <file>         Output file (default: stdout)
  --pretty                    Pretty-print JSON output
  --report                    Print conversion report to stderr

Example Output

A single Hevy workout converts to:

{
  "date": "2024-06-10T08:15:00.000Z",
  "name": "Upper Body",
  "durationSeconds": 5400,
  "exercises": [
    {
      "exercise": {
        "name": "Bench Press"
      },
      "supersetId": "1",
      "sets": [
        { "reps": 8, "weight": 80, "unit": "kg" },
        { "reps": 8, "weight": 80, "unit": "kg" },
        { "reps": 7, "weight": 80, "unit": "kg" }
      ]
    },
    {
      "exercise": {
        "name": "Bent Over Row"
      },
      "supersetId": "1",
      "sets": [
        { "reps": 8, "weight": 70, "unit": "kg" },
        { "reps": 8, "weight": 70, "unit": "kg" },
        { "reps": 8, "weight": 70, "unit": "kg" }
      ]
    }
  ]
}

Multiple workouts are output as a JSON array.

Troubleshooting

Date parsing issues

Hevy uses the format "10 Jun 2024, 08:15" in some exports and ISO timestamps in others. Both are handled automatically.

Exercises show as "unmapped"

This is normal. The converter maps ~60 common exercises to canonical names. Anything without a mapping is kept exactly as Hevy names it. Check the --report output to see which exercises were unmapped.