diff --git a/docs/source/robometer.mdx b/docs/source/robometer.mdx
index 3af822588..2d59eda90 100644
--- a/docs/source/robometer.mdx
+++ b/docs/source/robometer.mdx
@@ -78,7 +78,7 @@ The model predicts per-frame progress and success internally. The LeRobot reward
 from lerobot.rewards.robometer import RobometerConfig, RobometerRewardModel
 
 cfg = RobometerConfig(
-    pretrained_path="lilkm/Robometer-4B",
+    pretrained_path="lerobot/Robometer-4B",
     device="cuda",
     reward_output="progress",
 )
@@ -119,7 +119,7 @@ from lerobot.rewards import make_reward_model, make_reward_model_config, make_re
 
 cfg = make_reward_model_config(
     "robometer",
-    pretrained_path="lilkm/Robometer-4B",
+    pretrained_path="lerobot/Robometer-4B",
     device="cuda",
     image_key="observation.images.top",
 )
diff --git a/src/lerobot/rewards/robometer/compute_rabc_weights.py b/src/lerobot/rewards/robometer/compute_rabc_weights.py
index 3e985cb26..87d844340 100644
--- a/src/lerobot/rewards/robometer/compute_rabc_weights.py
+++ b/src/lerobot/rewards/robometer/compute_rabc_weights.py
@@ -33,13 +33,13 @@ Usage:
     # Dense per-frame progress for one episode
     python -m lerobot.rewards.robometer.compute_rabc_weights \\
         --dataset-repo-id lerobot/libero_10_image \\
-        --reward-model-path lilkm/Robometer-4B \\
+        --reward-model-path lerobot/Robometer-4B \\
         --episodes 0
 
     # All episodes with batching
     python -m lerobot.rewards.robometer.compute_rabc_weights \\
         --dataset-repo-id lerobot/libero_10_image \\
-        --reward-model-path lilkm/Robometer-4B \\
+        --reward-model-path lerobot/Robometer-4B \\
         --batch-size 16
 """
 
@@ -218,13 +218,13 @@ Examples:
     # Dense per-frame progress for one episode
     python -m lerobot.rewards.robometer.compute_rabc_weights \\
         --dataset-repo-id lerobot/libero_10_image \\
-        --reward-model-path lilkm/Robometer-4B \\
+        --reward-model-path lerobot/Robometer-4B \\
         --episodes 0
 
     # All episodes, smaller batches for memory-constrained GPUs
     python -m lerobot.rewards.robometer.compute_rabc_weights \\
         --dataset-repo-id lerobot/libero_10_image \\
-        --reward-model-path lilkm/Robometer-4B \\
+        --reward-model-path lerobot/Robometer-4B \\
         --batch-size 16
         """,
     )
diff --git a/src/lerobot/rewards/robometer/configuration_robometer.py b/src/lerobot/rewards/robometer/configuration_robometer.py
index 063c4a9be..fdaf7c9fd 100644
--- a/src/lerobot/rewards/robometer/configuration_robometer.py
+++ b/src/lerobot/rewards/robometer/configuration_robometer.py
@@ -50,7 +50,7 @@ ROBOMETER_SPECIAL_TOKENS = (
 class RobometerConfig(RewardModelConfig):
     """Configuration for the Robometer reward model."""
 
-    pretrained_path: str | None = "lilkm/Robometer-4B"
+    pretrained_path: str | None = "lerobot/Robometer-4B"
     image_key: str = OBS_IMAGES + ".top"
     task_key: str = "task"
     default_task: str | None = None