Your plan updates itself after every workout.

Kaizer recalculates the 1RM and adjusts weights and reps after every session your client logs — with real sets, RIR and NSCA tables. You always have the final say.

An adaptive workout app where each session updates the next routine.

The complete auto-progression loop — from the client closing a set to seeing the new weight in their next session.

Session A · Bench pressClosed

SetKgRepsRIR

16083

26082

36082

46082

Step 01

The client trains and logs.

Everything is recorded from the client app — no spreadsheets, no WhatsApp screenshots.

Sets, reps, weight and RIR per set
Under 30 seconds per session
Works offline at the gym

Kaizer engine320 ms

72 kg

Previous 1RM

76 kg

New 1RM

NSCA tableHigh confidence

Step 02

Kaizer recalculates the 1RM.

As soon as the client closes the session, the engine crosses real performance with reference tables.

Completed sets × reps × reported RIR
NSCA-derived tables
Ready 320 ms after the session closes

Next session · Mon 09:00+2.5 kg

Bench press

Weight6062.5 kg

Sets4 × 6-8

ApplyEdit

Step 03

The next session adjusts itself.

The new weight and reps are already in the routine when the client opens the app again.

Respects minimum increment and rep range
Automatic or suggestion — you decide
Every adjustment shows its reasoning

As a progressive overload app, the algorithm respects the minimum increment, rep range and progression mode (weight-driven, reps-driven or RIR-driven) you configured per exercise. Initial routine generation runs through an LLM that takes goals, preferences, limitations, equipment and experience as inputs. The coach retains full editorial control.

Monday's perfect program dies by Wednesday.

Your client slept badly, their knee acted up, they missed two sessions — and re-planning forty programs by hand doesn't scale. Kaizer adjusts the next session with every piece of feedback, inside the guardrails you set: the programming adapts itself, you approve.

The next session adjusts itself, inside your guardrails.

Start free

You decide the control level.

The AI proposes or applies, depending on how you configure it per client. Every change carries visible reasoning: why the weight went up or down.

Automatic or suggestionToggle on: it applies itself. Off: it queues as an editable proposal.
Visible reasoningEach adjustment explains the data behind it. Nothing is a black box.
Your rules, alwaysRespects rep range, minimum increment and progression mode per exercise.

Client · Sofía M.

Automatic progression

Bench press · next Monday+2.5 kg

4×8 @ RIR 2, two sessions in a row → 1RM 76 kg

ApplyPostpone

What the AI does not do today.

We prefer to do one thing well. This is explicitly out of scope — and it's written here so there are no surprises.

The coach-client relationship is what drives retention. The AI exists so you can keep that quality at scale, not to remove you from the equation.

Out of scope

Doesn't swap exercises on its ownIt only substitutes when the coach asks.

Doesn't message your clientEvery conversation is started by the coach.

Doesn't replace your judgmentToggle off, everything is an editable suggestion.

Doesn't sell dataClient data never goes to third parties.

How it compares to competitor 'AI'.

Trainerize and Harbiz generate initial routines and stop there. Kaizer is the only one that recalculates and adjusts after every session.

TrainerizeHarbizTrueCoachHevyGenerates the planAdjusts after every session1RM via NSCA tablesCoach as final editor

Trainerize and Harbiz generate initial routines with AI and stop there. TrueCoach and Hevy program by hand; Hevy ships the client app under the Hevy brand. Kaizer is the only one that recalculates the 1RM and adjusts weights and reps after every session, always keeping the coach as the final editor.

See full comparisons

No. And that's not the idea. The AI proposes (or applies, depending on your toggle) specific adjustments based on client data; you approve, edit, or let it run unattended. The coach-client relationship is what drives retention. The AI exists so the coach can maintain quality at scale — not to remove them from the equation.

Trainerize AI and Harbiz Genius generate initial routines with AI and stop there. The routine doesn't automatically change when client data changes. Kaizer recalculates the client's estimated 1RM after every workout they log (using sets, reps and RIR) and updates the target weights and reps of upcoming sessions.

Turn off the 'weightUpdatorEnabled' toggle for that client and changes stay queued as suggestions in the dashboard rather than applied. Each change comes with the reasoning (why the weight went up or down); you can edit it, defer it or discard. The system respects the rep range, the minimum increment and the progression mode you configured per exercise.

Yes. Adjustment kicks in as soon as the client logs their first session: with sets, reps and RIR from one session there's already a base for the next target. The first weeks the progression is more conservative (weights move in small increments respecting the minimum increment) and gets more precise as logged sessions pile up.

Yes, at generation time. You load per-client restrictions (no knee, no lumbar, pregnancy, etc.) and the LLM respects those rules when building the initial routine. The subsequent adjustment only modifies weights and reps within the exercises that are already in the routine — to swap an exercise for a new injury, you do the swap manually from the dashboard.

Customer reports show that in the first month coaches save 4–6 hours/week on manual programming. Improvements in client retention and adherence appear in the 2–4 month range once the AI has enough data to propose precise adjustments.

Yes on all paid plans (Starter, Pro, Unlimited, Gym). The Free plan has AI enabled with limits on re-evaluation frequency. Pricing on kaizer.app/en/pricing.

Explore other features.

Metrics

Your plan updates itself after every workout.