Lab-grade biomechanics.
Without the lab
Model Health turns simple smartphone videos into objective 3D movement data that medical staff and research teams can trust. Better decisions in rehab, performance, and R&D — without expensive labs or complex setups.
Model Health converts smartphone videos into
validated, decision-ready movement data.
Founded by Stanford researchers, our technology generates accurate 3D movement analysis from just smartpone videos.
Markerless motion capture, anywhere
Automated data analysis for most frequent physical tests
Health & performance dashboards and reports
Rehab and return-to-play
decisions
Track rehabilitation progress, identify compensation patterns before they become problems, and personalize protocols based on objective data.
Performance
monitoring
.webp)
Spot performance drops, identify technique improvements, and uncover potential. All grounded in individual baselines and normative datasets.
Research and product
development
Accelerate research cycles with automated processing, deployable in diverse environments where standard motion capture falls short.
Platform
integration
Integrate directly into your platform via SDK and access movement data programmatically to power your own workflows and products.
Built for professionals working with human movement. Different roles, different workflows. One shared need: data you can trust.
For clinical and medical staff who want to move beyond observation
Identify the root cause
of movement limitations with objective biomechanical data
Track rehabilitation progress
over time using consistent, repeatable measurements
Support clinical decisions
with clear, data-backed recommendations
Communicate progress
clearly to patients with visual reports they can understand
.webp)
For performance teams who want to quantify what drives athletic performance
Analyze movement strategy
not just load or output
Track asymmetries and coordination patterns
and how they evolve across the season
Support readiness and return-to-performance decisions
with objective data
Individualize training interventions
based on individual baselines and team benchmarks
.webp)
For researchers and engineers who want to scale biomechanics beyond the lab
Capture research-grade 3D kinematics
without marker-based systems
Run reproducible protocols
across sites and cohorts
Accelerate research and development cycles
through automated movement analyses completed in minutes
Export raw and processed data
for rigorous downstream analysis
.webp)
For product teams who want to build fast
Integrate 3D motion analysis in days
with a few lines of code, no computer vision and biomechanics teams required
Ship validated, standardized kinematics
inside your existing product workflows
Route results into your EMR or AMS automatically
without duplicate entry or file exports
Access data programmatically
for analytics pipelines, research workflows, and ML applications
Hear it from our customers
1 / 5
What our clients say:
I used to do running analysis with 2D video and software that required me to mark joints at specific moments in the gait cycle to estimate angles. I really appreciate being able to collect and visualize full kinematics time series with the same 2D video and built-in analysis with Model Health. I'm developing a new running analysis program using Model Health and I'm confident that the ability to provide more comprehensive data will set me apart and allow me to provide much better results to the runners I work with
Jessica Hunter
PhD, Kinesiology, Running Certified Coach
Model Health has been a huge boost to our small private practice. We have been able to implement it mainly with 3D running gait analyses as well as incorporate it into our ACL return to sport testing workflows. The set-up is quick, data collection is seamless, and the data processing continues to get faster. The reports that it generates are accurate, consistent, and are high quality for kinematic data, with easy-to-understand visualizations for joint angles and key metrics.
Alex Peters
Sports Physical Therapist at Impulse PT
Using this system in my clinic has been a game change. It provides clear, actionable insights into movement patterns, helping me make more informed decisions for my patients' treatment plans. The accuracy and ease of use make it an invaluable tool for improving patient outcomes.
Dr. Wilfredo Agredo
Medical Doctor
.svg){kind=logo}
Model Health's ability to quickly understand our needs, combined with their responsiveness, make our collaborations particularly effective!
Decathlon
R&D and QA developer at Decathlon
It's been great to implement Model Health in our clinic. It challenges our thoughts and gives us a way to objectively look at movements for those with challenging lingering pain when strength, Range Of Motion, and structure are "healthy". Especially useful for those working with athletes where we are unable to "see" high speed movements
Vien Vu
High-Performance Sports Physical Therapist at Stanford Athletic
We’re excited to integrate this innovative motion analysis into our ACL pathway at the POSM Performance Lab. This technology allows us to quantify contact time, RSI, and knee and hip joint kinematics during the triple hop test, providing a more comprehensive assessment of capacity following ACL injury.
Nic Blackah
Physiotherapist at POSM Performance Lab
Developed in academic research. Built for real-world use
Built on the foundations of OpenCap, Model Health delivers validated movement analysis for rehab, performance, and R&D.
Get started with Model Health
See how Model Health fits into your clinical, performance, or research workflow. Create your free account and start analyzing movement in minutes.
FAQs
This FAQ covers the questions we’re asked most often. Haven’t found what you are looking for? Our team is here to help.
Can I use Model Health outdoor?
Yes, you can use Model Health anywhere. Just ensure you have a reliable internet connection. You should be able to collect data under a slow network connection, although you'll probably have to wait a bit longer for the videos to upload.
How long does it take for an activity recording to be processed ?
It only takes two to five minutes from video upload to 3D model reconstruction to full movement analysis, incl. raw kinematics data, test-specific metrics and detailed PDF report of the activity. Duration mostly depends on the duration of the recording itself.
What equipment do I need to use Model Health?
You only need a computer or a tablet to launch recordings and analyze data, 2 iOS devices (any iPhone or iPad released in 2018 or later) mounted on 2 tripods, a printed checkerboard mounted on a flat surface for accurate set up calibration, and a reliable internet connection for best experience.
Is there a free trial?
Yes, you can request your 30-day free trial here
Where can I download the mobile app?
The Model Health iOS app is used to record videos and is controlled by our web app. The iOS app is available on the App Store and runs on both iPhone and iPad. Search for "Model Health" or use this direct link.
What browsers does the web app support?
Our web app is optimized for laptops and tablets. On laptop, the web app works best on Google Chrome (recommended), Firefox, Safari, and Edge.
On iPads, Safari is recommended. On Android tablets, Google Chrome is recommended.
Do I need a stable internet connection?
Yes. Both the iOS app and the web application require a stable internet connection to function properly. Our app supports asynchronous video upload, for situations where network is slow. Data will be automatically be processed as soon as videos get uploaded from each phone. Under very slow network conditions, your video recordings will be safely stored on device. Check the app as soon as you are back in an area with good network.
What types of movements can Model Health capture?
Model Health allows you to reconstruct precise 3D biomechanical models for just any movement. As long as the participant stays within the camera field of view, most functional movements can be captured.
Additionally, for a set of 12 movement types, we support automated analysis and key metric extraction. Movements include squat exercise, Counter Movement Jump, Drop Vertical Jump, hop tests, cutting maneuvers, 5-0-5 test, sit-to-stand transfer, overground and treadmill running and walking, and much more soon.
Most analyses automatically support and detect single and double leg variations, and any number of repetitions. For cutting maneuver, we support automatic detection of cut angles. For hop tests, we support broadjump, single, double and triple hop, without or without crossover.
What does the output look like?
Model Health produces whole-body 3D kinematics using a scientifically-validated musculoskeletal model with 36 degrees of freedom, covering joints from the ankle up through the hip, lumbar spine, shoulder, and elbow. The model is built using OpenSim, a widely used open-source platform in biomechanics research. The 3D model data can be visualized, alongside synchronized videos, in our web application.
The automated analysis outputs combine test metrics (such as contacts and completion time, velocities, key peak joint angles (incl. dynamic knee valgus for relevant tests), body orientation, and more). Data are presented as dashboard in our web applications and key be exported in various format, incl. fully automated PDF reports and csv files.
You can access a demo dashboard from the "See it in Action" button at the top of our website home page and download sample PDF reports for each activity in the App section of this website.
How accurate is it compared to traditional motion capture?
Model Health is built on markerless technology, eliminating the need for physical markers while still generating biomechanically meaningful joint kinematics. The technology builds on OpenCap, developed by our founders while at Stanford University. For detailed accuracy benchmarks, consult our validation papers: first article and second one.