The research lab for climbing

Climbing processes such as routesetting, figuring out beta or even classifying the difficulty of climbing routes are challenging and notoriously subjective tasks, even for experienced climbers. We are a R&D-oriented startup which brings innovative tools in order to make climbing a systematic discipline.


Our research angles

to solve climbing

Grade System

The current scale is based on a comparison sort: V10 is harder than V9 and easier than V11. However the relation "to be harder/easier than" is highly subjective. Climbing needs a real scale that is objective and systematic in order to quantify performance and enhance the progression of the disciple.


Beta is the technical guidance for a mouvement sequence through a problem. It's the strategy aspect of the sport. To lead climbing to the next step, we need to theorize beta to improve athletes' sequences and decision making.

Route setting

The continuous route creation process is in the blood of climbing. Though, it requires high skills and adjustment to tests the route multiple times, resulting in route setter/route enablers being exposed to injuries. Making route setting less dangerous, accessible, and affordable will bolster the expansion of climbing around the world.


Climbing is highly based on experience. Machine Learning is the science which permits to learn from data, that means learning from the user experience. At DeepClimb we believe that Artificial Intelligence, thanks to machine learning, can expand our knowledge on our sport.

Our first platform,
7th Face:
The AI-powered Spray Wall App

Add, share and for the first time, analyse boulder and beta form spraywall.
Thanks to Deep Learning, our algorithm automatically grades boulders.
Our AI is able to analyse and figure out the gesture repertoire of each climber in order to personalise their search page

Convinced to play a major role in the climbing history?

Loading Page.png
View Boulder - Beta Version@3x.png


Create and share your own spray wall problems


The SmartSearch algorithm allows you to find always the boulder that suits you the most


The AI grades problems for you: No more sandbagging, only objective grades!

Capture d’écran 2020-03-04 à 11.54.29.png


DeepClimb is still in development.
We will provide periodic updates when we cross new milestones. If you’re interested in chatting with us we’d love to talk.