Supported Models

• Robometer (https://robometer.github.io)
• SOLE-R1 (https://philipmit.github.io/sole-r1/)
• TOPReward (https://topreward.github.io/webpage/)
• RoboReward (https://arxiv.org/abs/2601.00675)
• OpenAI models (e.g., "GPT-5")
• Google models (e.g., "Gemini-3-Pro-Preview")

ToDos

• Enable fine-tuning of reward models on custom datasets

📦 File Structure

roboreason/
├── roboreason/         # Main package
│   ├── robometer/         # Robometer code
│   ├── sole.py            # SOLE-R1 code
│   ├── roboreward.py      # RoboReward code
│   ├── topreward.py       # TOPReward code
│   └── api_models.py      # OpenAI and Gemini APIs
├── test_videos/        # Example videos to test
├── model_outputs/      # Example videos showing model outputs
├── docs/   
│   ├── lerobot_dataset_reward_annotation.mdx  # Examples showing integration with lerobot datasets
└── pyproject.toml      # Dependencies (uv)

Install

Option 1: quick pip install

pip install -U roboreason

Option 2: use uv for dependency management

1. Clone the repository:

git clone https://github.com/philipmit/roboreason

2. Install uv

pip install uv

3. Sync environment

uv sync

4. Activate environment

source .venv/bin/activate

Optional: Pre-download model checkpoints

# SOLE-R1 (8B) 
python -c "from roboreason.utils.model_utils import get_model_dir; get_model_dir('sole-r1')"

# Robometer (4B)
python -c "from roboreason.utils.model_utils import get_model_dir; get_model_dir('robometer')"

# TOPReward (based on Qwen3-VL-8B)
python -c "from roboreason.utils.model_utils import get_model_dir; get_model_dir('topreward')"

# RoboReward (8B)
python -c "from roboreason.utils.model_utils import get_model_dir; get_model_dir('roboreward')"

> **Note:** Robometer is ~8GB. SOLE-R1, RoboReward, and TOPReward are ~17GB each.

Optional: Download all test videos and example model outputs from google cloud

# 1) Install gcloud: https://cloud.google.com/sdk/docs/install

# 2) Go to target directory
# cd /path/to/roboreason

# Optional: disable credentials so you don't have to authenticate
gcloud config set auth/disable_credentials True

# Download test videos
gcloud storage cp --recursive gs://roboreason-view-videos-philip/test_videos ./

# Download example model outputs
gcloud storage cp --recursive gs://roboreason-view-videos-philip/model_outputs ./

# Optional: re-enable credentials afterward if you disabled them above.
gcloud config set auth/disable_credentials False

Quick start: Example reward generation and plotting

# pip install -U roboreason
import roboreason as rr

video_paths = ['test_videos/robosuite/lift/unsuccessful/robosuite_lift_episode_11_unsuccessful_max_reward_37.mp4']
task_description="Pick up the cube from the table."

# Robometer
rewards, success_probs = rr.generate(model="Robometer",  task_description=task_description, video_paths=video_paths, view_type_per_video=['external'], verbose=False)
output_robometer = {"model": "Robometer", "rewards": rewards[0]}

# SOLE-R1
rewards, reasoning_traces = rr.generate(model="SOLE-R1",  task_description=task_description, video_paths=video_paths, view_type_per_video=['external and wrist'], verbose=False)
output_sole = {"model": "SOLE-R1", "rewards": rewards[0], "reasoning_traces": reasoning_traces[0]}

# Optional: Ground-truth rewards (available for test videos from sim environments)
import json
with open(video_paths[0].replace(".mp4", "/data.json"), 'r') as f:
    data = json.load(f)

output_groundtruth = {"model": "Ground truth", "rewards": data['ground-truth rewards']}

# Plot
rr.video_plot(outputs=[output_groundtruth, output_sole, output_robometer], plot_save_path='model_outputs/combined/robosuite/lift/unsuccessful/robosuite_lift_episode_11_unsuccessful_max_reward_37.mp4', video_path = video_paths[0])

Examples for generating across all models

Robometer

import roboreason as rr

rewards, success_probs = rr.generate(
    model="Robometer",  
    task_description="Pick up the cube from the table.", 
    video_paths=['test_videos/robosuite/lift/unsuccessful/robosuite_lift_episode_11_unsuccessful_max_reward_37.mp4'], 
    view_type_per_video=['external'],
    verbose=False
)

SOLE-R1

import roboreason as rr

rewards, reasoning_traces = rr.generate(
    model="SOLE-R1",  
    task_description="Pick up the cube from the table.", 
    video_paths=['test_videos/robosuite/lift/unsuccessful/robosuite_lift_episode_11_unsuccessful_max_reward_37.mp4'], 
    view_type_per_video=['external and wrist'],
    verbose=False
)

TOPReward

import roboreason as rr

rewards = rr.generate(
    model="TOPReward",  
    task_description="Pick up the cube from the table.", 
    video_paths=['test_videos/robosuite/lift/unsuccessful/robosuite_lift_episode_11_unsuccessful_max_reward_37.mp4'], 
    view_type_per_video=['external'],
    verbose=False
)

RoboReward

import roboreason as rr

rewards = rr.generate(
    model="RoboReward",  
    task_description="Pick up the cube from the table.", 
    video_paths=['test_videos/robosuite/lift/unsuccessful/robosuite_lift_episode_11_unsuccessful_max_reward_37.mp4'], 
    view_type_per_video=['external'],
    verbose=False
)

GPT-5 (and other OpenAI models)

import roboreason as rr

# requires OpenAI API key: https://developers.openai.com/api/docs/quickstart
API_KEY = "..."

rewards, reasoning_traces = rr.generate(
    model="GPT-5",  
    task_description="Pick up the cube from the table.", 
    video_paths=['test_videos/robosuite/lift/unsuccessful/robosuite_lift_episode_11_unsuccessful_max_reward_37.mp4'], 
    view_type_per_video=['external'], 
    key=API_KEY, 
    verbose=False
)

Gemini-3-Pro (and other Google models)

import roboreason as rr

# requires Gemini API key: https://ai.google.dev/gemini-api/docs/api-key
API_KEY = "..."

rewards, reasoning_traces = rr.generate(
    model="Gemini-3-Pro-Preview",  
    task_description="Pick up the cube from the table.", 
    video_paths=['test_videos/robosuite/lift/unsuccessful/robosuite_lift_episode_11_unsuccessful_max_reward_37.mp4'], 
    view_type_per_video=['external'], 
    key=API_KEY,
    verbose=False
)

Video plotting

import roboreason as rr

# Robometer
rewards, success_probs = rr.generate(model="Robometer",  task_description=task_description, video_paths=video_paths, view_type_per_video=['external'])
output_robometer = {"model": "Robometer", "rewards": rewards[0]}

# SOLE-R1
rewards, reasoning_traces = rr.generate(model="SOLE-R1",  task_description=task_description, video_paths=video_paths, view_type_per_video=['external and wrist'])
output_sole = {"model": "SOLE-R1", "rewards": rewards[0], "reasoning_traces": reasoning_traces[0]}

# Optional: Ground-truth rewards (available for test videos from sim environments)
import json
with open(video_paths[0].replace(".mp4", "/data.json"), 'r') as f:
    data = json.load(f)

output_groundtruth = {"model": "Ground truth", "rewards": data['ground-truth rewards']}

rr.video_plot(
    outputs=[output_sole, output_robometer], 
    plot_save_path='model_outputs/combined/robosuite/lift/unsuccessful/robosuite_lift_episode_11_unsuccessful_max_reward_37.mp4', 
    video_path = 'test_videos/robosuite/lift/unsuccessful/robosuite_lift_episode_11_unsuccessful_max_reward_37.mp4',
    verbose=False
)

Inference and plotting across multiple videos

import roboreason as rr
import glob
import json

video_paths = glob.glob('test_videos/robosuite/lift/unsuccessful/*')

## INFERENCE

# Robometer for all videos
rewards_robometer, success_probs_robometer = rr.generate(model="Robometer",  task_description=task_description, video_paths=video_paths, view_type_per_video=['external'])
# SOLE-R1 for all videos
rewards_sole, reasoning_traces_sole = rr.generate(model="SOLE-R1",  task_description=task_description, video_paths=video_paths, view_type_per_video=['external and wrist'])


## PLOTTING
plot_save_dir = 'model_outputs/'
for video_idx in range(len(video_paths)):
    output_robometer = {"model": "Robometer", "rewards": rewards_robometer[video_idx]}
    output_sole = {"model": "SOLE-R1", "rewards": rewards_sole[video_idx]}
    # Optional: Ground-truth rewards (available for test videos from sim environments)
    with open(video_paths[0].replace(".mp4", "/data.json"), 'r') as f:
        data = json.load(f)
    
    output_groundtruth = {"model": "Ground truth", "rewards": data['ground-truth rewards']}
    rr.video_plot(
        outputs = [output_sole, output_robometer], 
        plot_save_path = plot_save_dir + video_paths[video_idx].split('test_videos/')[-1] , 
        video_path = video_paths[video_idx],
        verbose = False
    )

rr.generate

rr.video_plot

Acknowledgements

RoboReason builds upon the following repos:

• RewardScope (https://github.com/philfung/reward-scope)
• Robometer (https://github.com/robometer/robometer)
• TOPReward (https://github.com/TOPReward/TOPReward)