Lifter or coach?I'm a lifter →·I coach athletes →

Most apps count your sets.
We weigh them.

A training log that scores every rep by how close it landed to failure — then shows you what your month actually produced, muscle by muscle.

Train alone or with a coach · start on web now, mobile app coming soon

Set credit modellive
10
8.0· 2 RIR
Counts as
0.60
of one hard set
Fatigue cost
0.73
paid for that credit

One or two in reserve. This is where most productive training lives — 60% of a hard set.

Move the rep slider: at the same RIR the credit barely changes. That is deliberate. When sets are taken equally close to failure, hypertrophy per set looks similar across roughly 5 to 30 reps — so proximity to failure drives the number, not rep count.

This is a model, not a measurement. It weighs every rep by how close it was to failure on a continuous curve — there is no cut-off after which reps stop counting.

First things first

It's a training log.
That part just works.

Everything above is what happens after you log. The logging itself is fast, offline-tolerant, and free.

Sets, reps, weight, RPE
Global exercise library + your own
Build a routine once, reuse it
Warm-ups, drop sets, myo-reps, supersets
Personal records and estimated 1RM
History calendar and streaks
Keeps your session if the signal drops
Free tier covers all of this
The result that started this

Same RPE, wildly different rep counts,
nearly identical stimulus.

Take a set to RPE 8 — two reps left in the tank. Do five. Do ten. Do twenty. The work looks completely different in your logbook. The model scores it almost the same.

SetRepsRPERep-equivalentsSet credit
A58.02.860.57
B108.03.010.60
C208.02.980.60
D105.01.030.21
E1010.05.001.00

Sets A–C: rep count moves, stimulus barely does. Sets B, D, E: rep count is fixed, proximity to failure moves everything. These are the model's real outputs, not illustrations.

Why we built it

Two lifters do twenty sets of chest in a week. One grows, one stalls. Set count didn't tell them apart — everything the set count throws away did.

Where the set landed. Whether the chest was doing the work or the triceps were. Whether the last three weeks were a ramp or a plateau. Every training log records the first number and quietly discards the rest.

So we kept the rest. Not as a score out of a hundred — as the actual quantities, with the assumptions written down and the limits stated.

What it measures

Six numbers your logbook is throwing away.

Set credit

Every rep is weighted by its distance from failure on a continuous curve — no arbitrary "effective reps start at 5 RIR" cutoff.

RPE 5 0.21 set
RPE 7 0.44 set
RPE 8 0.60 set
RPE 9 0.79 set
RPE 10 1.00 set

Direct vs indirect volume

Working as a prime mover and working as a helper are not the same stimulus, so they are never added together.

Lats38%
Erectors26%
Mid-traps19%
DirectIndirect

Stimulus ROI

Every session buys stimulus and pays in fatigue. This is the exchange rate.

stimulus 54.4
fatigue 65.4
───────────────
ROI 0.83 efficient

Training load

Last 7 days against your last 28. A workload signal — not a medical one.

<0.8
0.8–1.3
1.3–1.5
>1.5

Currently 1.06 — inside the band you want.

Muscle map

Front and back, broken down to the sub-muscle. Lats and erectors are not "back".

BACK ████████ 31%
lats ████ 38%
erectors ███ 26%
mid-traps ██ 19%

And the boring parts done properly

Build a routine once and run it as many times as you want, or start empty and decide as you go. Warm-ups don't pollute your volume. Drop sets, myo-reps and supersets are modelled the way you actually perform them, not flattened into "a set". Finish one routine and the next queues up — skip it or push it back, your call. Personal records, estimated 1RM progression, and a history calendar that shows consistency without nagging you about it.

If you coach

Your whole roster's load,
on one screen.

Coaching ten athletes means ten sets of numbers you can't hold in your head. The roster surfaces the ones that moved — who ramped too fast, who went quiet, who's ready for more.

Athletes · 8 activeLoad ratio
AthleteLast sessionACWR
Deniz A.Today1.42
Mert K.Yesterday1.08
Selin T.2 days ago0.94
Baran Y.9 days ago0.61
One athlete, full drilldown. The same numbers they see — load ratio, stimulus ROI, effective volume, wellness trends, body measurements — laid out for a coach reading five athletes in ten minutes.
Build a program, assign it, move on. Blocks, days, exercises, prescribed sets with target RPE. Assign it to one athlete or several; their app picks it up the same day.
An agenda that isn't a WhatsApp thread. Sessions, assignments and what each athlete is due to do next, in one calendar.
A shared exercise library. Your own exercises, your own muscle splits, applied across every athlete you coach.
Messages where the data is. Notes and replies sit next to the session they're about, not three apps away.

Athletes keep their own account either way. A coached athlete gets their program from you; an independent one runs their own routines. Same app, same numbers.

Read the full coach walkthrough →
Method & limits

What this model does not do.

Fitness tech has a credibility problem because everyone overclaims. Here are our limits, stated before you ask for them.

It does not

  • Measure hypertrophy. It estimates stimulus.
  • Predict or prevent injury. Load ratios are a workload signal, not a medical one.
  • Measure muscle damage. Fatigue cost is a proxy, and it is labelled as one.
  • Claim a patented algorithm. The references are listed below.

It does

  • Weight every rep continuously by proximity to failure.
  • Separate direct from indirect muscle contribution, per exercise.
  • Hold stimulus and its fatigue cost as two numbers, never merged into one score.
  • Let you edit any exercise's muscle split yourself.
Core references behind the stimulus model:
  • Robinson et al. 2024 — dose–response between estimated RIR and hypertrophy
  • Refalo et al. 2023 — proximity-to-failure meta-analysis
  • Fisher, Steele & Smith 2022 — failure as a continuous, not binary, variable
  • Gabbett 2016 — acute:chronic workload framework
  • + 25 peer-reviewed sources behind the exercise muscle-contribution dataset
Pricing

Start free. Turn on the scientific layer whenever you want.

Sign-up is free, no card required. Turn on scientific analytics — ACWR, Stimulus ROI, effective volume — whenever you're ready.

Individual Free
Free

Personal training, routines and basic tracking.

  • Workout logging
  • Personal routines
  • Global exercise library
  • Basic analytics
  • Personal records
Sign up free →