Skip to content
JAX EvoGym

Quick Start

Create a world

Section titled “Create a world”
import numpy as np
from jax_evogym import EvoWorld, H_ACT, SOFT, V_ACT


world = EvoWorld()
world.add_from_array("robot", np.array([[H_ACT, SOFT, V_ACT]]), 0, 0)
world.to_json("robot.json")

EvoWorld is a numpy-only world definition. No JAX required at this stage.

Run an environment

Section titled “Run an environment”
import jax.numpy as jnp
import numpy as np
from jax_evogym import H_ACT, SOFT, V_ACT, WalkerV0


body = np.array([[H_ACT, SOFT, V_ACT]])
env = WalkerV0(body)


obs, state = env.reset()
action = jnp.ones(env.n_actuators)
obs, state, reward, done = env.step(state, action)

Environments handle the full build pipeline internally: compile the world template, instantiate the built world, and expose a JIT-compiled step function.

Use the builder API directly

Section titled “Use the builder API directly”
import numpy as np
from jax_evogym import EvoWorld, FIXED, H_ACT, SOFT, V_ACT
from jax_evogym import compile_world_template, instantiate_world


world = EvoWorld()
world.add_from_array("ground", np.array([[FIXED, FIXED, FIXED, FIXED]]), 0, 0)
world.add_from_array("robot", np.array([[H_ACT, SOFT, V_ACT]]), 1, 2)


template = compile_world_template(world, robot_name="robot")
built = instantiate_world(template)


print(built.sim_state.positions.shape)
print(built.static_collider_data.point_positions.shape)

The builder API gives you direct access to SimState, CollisionData, ActuatorInfo, and all other simulation primitives. Use this when you need control beyond what the built-in environments provide.

Where to go next

Section titled “Where to go next”