# chungmin99/pyroki

A Modular Toolkit for Robot Kinematic Optimization

Repository: https://github.com/chungmin99/pyroki
Canonical: https://ross.abutalabs.com/products/pyroki
Homepage: https://pyroki-toolkit.github.io/
Language: Python
License: MIT
License Family: permissive
Last push: 2026-04-01T00:13:52+00:00

## Health v2 (maintenance only)
Score: 53/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 75, release rhythm 35, longevity 34
- inputs: {"age_days": 484, "days_push": 155, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1682, forks 183 (observed 2026-08-28T04:05:21.719980+00:00)

## What it is
PyRoki is a modular Python toolkit for robot kinematic optimization built on JAX, providing differentiable forward kinematics from URDFs, collision primitives, and composable costs. It unifies inverse kinematics, trajectory optimization, and motion retargeting via a Levenberg-Marquardt solver with manifold optimization and hard constraints, running on CPU, GPU, and TPU.

## Use cases
- solve inverse kinematics for a robot arm from a URDF
- optimize robot trajectories with collision avoidance
- retarget human hand or body motion to robot hands and humanoids
- compute differentiable forward kinematics in a JAX pipeline
- optimize end-effector pose and manipulability objectives
- generate collision capsules from robot meshes automatically

## When to choose
- you need differentiable or batched kinematics and IK in Python with GPU/TPU support
- you want composable costs for IK, trajectory optimization, or motion retargeting in one framework
- your robot uses revolute, continuous, prismatic, or fixed joints in a kinematic tree

## When to avoid
- you need sampling-based motion planning with graphs or trees
- you require closed-loop kinematic chains or parallel manipulators
- you need fast mesh-level collision checking in collision-heavy scenarios

## Facets
- artifact type: library
- maturity: active
- function: simulation, math, machine-learning
- domain: robotics, gpu-computing
- platform: python, cross-platform
- tags: robot-kinematics, inverse-kinematics, jax, urdf, trajectory-optimization, motion-retargeting, collision-avoidance, differentiable-programming, algorithms, gpu

## Member repositories
- chungmin99/pyroki (main) score 53

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:05:21.719980+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T03:41:21.747197+00:00, confidence not recorded.
  - readme: https://github.com/chungmin99/pyroki (fetched 2026-08-28T04:05:21.719980+00:00, sha 902b2b2ea934)
  - homepage: https://pyroki-toolkit.github.io/ (fetched 2026-08-29T11:14:16.741882+00:00, sha 8a610126af31)
- Data as of 2026-08-30T08:39:29.467469+00:00.