Muse MuseConcept

Independent concept built on Meta's Model API. Not an official Meta product.

Gym Snap Coach

Status: Live.

  1. Your muse

    Take an image of a gym, segment, crop, create an image of how to use a machine and a workout routing

  2. Done. Gym Snap Coach is built and verified. npm run verify exits 0 with all six steps passing: typecheck, build, start, console, smoke (6 tests), and manifest.

  3. Failed

    Checks found problems

    The final routine is present, but its warm up and cool down sections show placeholder x and y bullets instead of real steps.

  4. Failed

    Checks found problems, try 2

    Smoke test

  5. Done. Fixed. npm run verify exits 0, all six steps passing, 6 of 6 tests green.

  6. Failed

    Checks found problems, try 3

    The final routine lists rowing machine twice and uses a full sentence as the fourth machine name.

  7. Done

    Checks passed, try 4

  8. The plan: Gym Snap Coach

    Photograph any gym, learn each machine, and get a routine from what is there.

    1. Upload one clear photo of the gym and enter your goal, level, and time available
    2. The app segments the photo and shows labeled boxes around each machine it finds
    3. Tap a machine to see its crop, what it works, setup and form steps, safety notes, and a...
    4. Tap Build My Routine to get an ordered workout using only the machines found in the photo
    See the full plan

    Who it's for

    People training in an unfamiliar gym, hotel gym, or home gym who do not know what the machines are or how to use them safely.

    How it works

    1. Upload one clear photo of the gym and enter your goal, level, and time available.
    2. The app segments the photo and shows labeled boxes around each machine it finds.
    3. Tap a machine to see its crop, what it works, setup and form steps, safety notes, and a generated how-to illustration.
    4. Tap Build My Routine to get an ordered workout using only the machines found in the photo.
    5. Follow the plan in order and check off exercises as you go.

    Screens

    • Upload and Detect. Upload a gym photo, enter goal, level, and time, and see labeled boxes over each detected machine.
    • Machine Guide. Show the cropped machine, muscles worked, setup and form steps, safety notes, and a generated how-to illustration.
    • Workout Routine. Show an ordered routine built from the detected machines, with sets, reps, rest, and a check-off list.

    What goes in

    • Image: One photo of a gym or workout area, JPG or PNG. Keep it under 4.5 MB. A wide, well-lit shot showing whole machines works best.
    • Text: Goal such as strength, fitness, or weight loss, experience level, minutes available, and any limits or injuries to avoid. Short plain text.

    What comes out

    • Image: Generated illustration of correct setup and form for the selected machine, with no real people or brands.
    • Text: Labeled equipment list, step-by-step machine instructions with form and safety cues, and an ordered workout routine with sets, reps, and rest times. Saved only in the browser.

    Models

    ModelUsed forWhy
    SAM 3.1Find and outline each machine in the gym photo so the app can crop and label individual equipment.Segmentation gives accurate boxes and crops. A vision model alone would guess locations.
    Muse WonderIdentify each cropped machine, explain how to use it, and build a workout routine from the machines found.It reads the crop images and writes clear instructions and a plan matched to goal, level, and time.
    Muse ImageCreate a clean illustration showing correct setup and form on the selected machine.A generated visual makes form cues easier to follow than text alone.

    Limits

    • Machine identification from a photo can be wrong, especially for similar cable or selectorized machines. Check the name and weight stack on the machine itself before use.
    • Generated illustrations show general form only. They are not a substitute for a qualified trainer, and joint angles or grips may not match your exact machine model.
    • Only equipment clearly visible in the photo is included. Machines hidden, cut off, or in shadow may be missed or cropped poorly.
    • The routine is general fitness guidance, not medical advice. Stop if you feel pain and talk to a professional about injuries or health conditions.
    • One routine and one illustration are generated per run to keep use fast and low cost.

    From your inspo

    The person gave a short text note only. They want a tool that takes a photo of a gym, segments and crops the machines in it, shows how to use a machine, and creates a workout routine from the equipment found.

    Takes

    • The core job: turn one gym photo into machine guidance and a workout routine.
    • The flow: detect equipment first, then explain one machine, then build a plan from what is available.
    • A practical and encouraging tone for beginners who feel unsure in a gym.

    Cost per run: about $0.05 (est.). SAM segmentation $0.0025 for one photo, plus Wonder identification and routine about 4,000 input tokens at $4 per million and 1,200 output tokens at $20 per million for about $0.040, plus one generated illustration at $0.01. Total about $0.05 per run.

  9. Recorded demo
  10. Done. Fixed. npm run verify exits 0 with all six steps passing and 6 of 6 tests green.

    40 m 17 s · Input 14.4M · Output 97.4k · $8.58 est.