AI route-setting instrument

See the beta before you pull on.

Upload a few photos of a boulder. Boulderyze reconstructs the wall’s 3D structure, maps every hold and volume, separates routes by colour, and reads you the best sequence of moves — drawn step-by-step on the photo and in plain words.

1–3
photos in
~45s
to full beta
3D
orbit-able z-map
Z-MAP · DEPTH ANYTHING V2 CAVE OVERHANG
12345 V6 · 6C+
LH RH LF RF
5 MOVES

STEREO + SHADING

3D wall & z-map

Multi-view parallax and shading infer a relative depth for every hold and volume — then render an orbit-able surface.

COLOUR SEGMENTATION

Routes by colour

Holds are grouped into problems by colour, with bolted-on volumes detected and outlined separately.

VISION LLM

Best beta, explained

A vision model ranks every candidate sequence and walks you through each hand & foot move, in order.

The workflow

Photo to beta in three moves

Try it now →
01

Upload

Drop 1–3 photos of the wall. We resize on-device and overlay a coordinate grid.

02

Reconstruct

We map holds, separate routes by colour, and infer the wall’s depth into a z-map.

03

Send

Review the detected holds, then get the ranked beta drawn move-by-move on the photo.

Stop guessing the sequence.

Boulderyze turns a phone photo into a coach. It’s free to start — bring your project.

Get started — it’s free
Boulderyze · AI boulder route & beta analysis
READING ROCK AS TERRAIN · v0.1