Model comparison

new_reward_dataset_irrh6c0s vs o3

Capability radar, value bars, and side-by-side posture.

← Back to Compare

new_reward_dataset_irrh6c0s
Not scored
value score
o3
93
value score
new_reward_dataset_irrh6c0s ctx
o3 ctx
200K

Capability radar

Value · context · multimodal · openness · speed posture

Value head-to-head

  • o393

Attribute tape

Fieldnew_reward_dataset_irrh6c0so3
DeveloperstrawberryroboticsincOpenAI
Context200K
ModalitiesText, Code
OpennessProprietary Model
Speed
Price$10/1M in · $40/1M out
Value score93
SummaryA proprietary robotics reward dataset published by Strawberry Robotics, Inc. on Hugging Face, designed to support reward-model training and evaluation for robot task performance. The dataset contains trajectory, video, or observation data paired with reward labels, following common embodied-AI reward-modeling pipelines.OpenAI reasoning model focused on hard problems.
Open new_reward_dataset_irrh6c0sOpen o3