Dataset Viewer
Auto-converted to Parquet
unique_data_id
stringlengths
25
25
taskId
stringlengths
25
25
task_name
stringlengths
22
386
category
stringclasses
2 values
subCategory
listlengths
1
1
application_website
stringclasses
409 values
tags
listlengths
0
7
benchmark
stringclasses
5 values
appType
stringclasses
2 values
difficulty
stringclasses
3 values
os
stringclasses
3 values
requires_login
stringclasses
3 values
videoSize
stringlengths
0
9
completedAt
stringdate
2025-04-08 05:56:26
2025-09-26 10:08:18
reasoning_steps
listlengths
0
25
metadata
stringlengths
2
612
screenshots
images listlengths
0
52
video_file
stringlengths
36
36
dom_snaps_file
stringlengths
0
39
events
stringlengths
0
662k
cmcc8u6yc00va1p1ydsdu52zy
cmcc8u68b00511p1yb12ej20b
"Use the top-right search bar to look up the term β€œprivacy policy,” open the official Privacy Po(...TRUNCATED)
BROWSER_TASK
[ "Search & Research" ]
FoxNews
["Information Retrieval & Analysis:Search","Navigation & Workflow Control:Navigation","Information R(...TRUNCATED)
Proprietary
SINGLE_APP
EASY
CROSS_PLATFORM
no
2025-07-11 15:06:45.223
[]
"{\"id\": 264800830346283, \"node\": \"DESKTOP-GFL3B1N\", \"model\": \"20FXS1BP00\", \"system\": \"W(...TRUNCATED)
[{"src":"https://datasets-server.huggingface.co/assets/anaisleila/computer-use-data-psai/--/{dataset(...TRUNCATED)
videos/cmcc8u6yc00va1p1ydsdu52zy.mp4
dom_snaps/cmcc8u6yc00va1p1ydsdu52zy.zip
"[{\"time_stamp\": 1245483.5058041, \"action\": \"move\", \"x\": 1467, \"y\": 47}, {\"time_stamp\": (...TRUNCATED)
cmcc8u6yc00v91p1yw2eruz95
cmcc8u68b00501p1y3zon7x4z
"From the FoxNews.com homepage, navigate to the β€œAbout” or β€œCorporate Information” page and (...TRUNCATED)
BROWSER_TASK
[ "Search & Research" ]
FoxNews
[ "Navigation & Workflow Control:Navigation", "Information Retrieval & Analysis:Extraction" ]
Proprietary
SINGLE_APP
EASY
CROSS_PLATFORM
no
2025-07-11 15:00:37.703
[]
"{\"id\": 264800830346283, \"node\": \"DESKTOP-GFL3B1N\", \"model\": \"20FXS1BP00\", \"system\": \"W(...TRUNCATED)
[{"src":"https://datasets-server.huggingface.co/assets/anaisleila/computer-use-data-psai/--/{dataset(...TRUNCATED)
videos/cmcc8u6yc00v91p1yw2eruz95.mp4
dom_snaps/cmcc8u6yc00v91p1yw2eruz95.zip
"[{\"time_stamp\": 1245169.1857704, \"action\": \"move\", \"x\": 1593, \"y\": 90}, {\"time_stamp\": (...TRUNCATED)
cmcc8u6yc00vm1p1yhjl1u0bf
cmcc8u68b005d1p1y3znjf5gc
"In the site footer, locate the β€œContact Us” or β€œFox News Careers” link. Open it and note tw(...TRUNCATED)
BROWSER_TASK
[ "Search & Research" ]
FoxNews
[ "Navigation & Workflow Control:Navigation", "Information Retrieval & Analysis:Extraction" ]
Proprietary
SINGLE_APP
EASY
CROSS_PLATFORM
no
2025-07-11 14:50:28.882
[]
"{\"id\": 264800830346283, \"node\": \"DESKTOP-GFL3B1N\", \"model\": \"20FXS1BP00\", \"system\": \"W(...TRUNCATED)
[{"src":"https://datasets-server.huggingface.co/assets/anaisleila/computer-use-data-psai/--/{dataset(...TRUNCATED)
videos/cmcc8u6yc00vm1p1yhjl1u0bf.mp4
dom_snaps/cmcc8u6yc00vm1p1yhjl1u0bf.zip
"[{\"time_stamp\": 1244566.5383122, \"action\": \"move\", \"x\": 1488, \"y\": 77}, {\"time_stamp\": (...TRUNCATED)
cmcc8u6yc00vi1p1y10rtsjn1
cmcc8u68b00591p1yi1575z0x
"Click β€œHealth” in the main navigation, select a nutrition-focused article, and summarize the ke(...TRUNCATED)
BROWSER_TASK
[ "News & Media" ]
FoxNews
[ "Navigation & Workflow Control:Navigation", "Information Retrieval & Analysis:Research" ]
Proprietary
SINGLE_APP
EASY
CROSS_PLATFORM
no
2025-07-11 14:28:14.710
[]
"{\"id\": 264800830346283, \"node\": \"DESKTOP-GFL3B1N\", \"model\": \"20FXS1BP00\", \"system\": \"W(...TRUNCATED)
[{"src":"https://datasets-server.huggingface.co/assets/anaisleila/computer-use-data-psai/--/{dataset(...TRUNCATED)
videos/cmcc8u6yc00vi1p1y10rtsjn1.mp4
dom_snaps/cmcc8u6yc00vi1p1y10rtsjn1.zip
"[{\"time_stamp\": 1243026.75937, \"action\": \"window_focus\", \"app_name\": \"chrome.exe\", \"wind(...TRUNCATED)
cmcc8u6yc00vf1p1yvn1ufxj2
cmcc8u68b00561p1yrreiu5vv
"Navigate to the β€œPodcasts” link in the site’s main or footer navigation, then list the distin(...TRUNCATED)
BROWSER_TASK
[ "News & Media" ]
FoxNews
[ "Navigation & Workflow Control:Navigation", "Information Retrieval & Analysis:Extraction" ]
Proprietary
SINGLE_APP
EASY
CROSS_PLATFORM
no
2025-07-11 14:03:30.332
[]
"{\"id\": 264800830346283, \"node\": \"DESKTOP-GFL3B1N\", \"model\": \"20FXS1BP00\", \"system\": \"W(...TRUNCATED)
[{"src":"https://datasets-server.huggingface.co/assets/anaisleila/computer-use-data-psai/--/{dataset(...TRUNCATED)
videos/cmcc8u6yc00vf1p1yvn1ufxj2.mp4
dom_snaps/cmcc8u6yc00vf1p1yvn1ufxj2.zip
"[{\"time_stamp\": 1241459.2144113, \"action\": \"move\", \"x\": 1490, \"y\": 72}, {\"time_stamp\": (...TRUNCATED)
cmcc8u6ym018l1p1yxhf18gc2
cmcc8u68o00ic1p1y9je7qdrf
"From the Templates section, open the β€œAll Templates” gallery, filter by the β€œPhotography” c(...TRUNCATED)
BROWSER_TASK
[ "Utilities & Tools" ]
Squarespace
["Navigation & Workflow Control:Navigation","Information Retrieval & Analysis:Filtering","Informatio(...TRUNCATED)
Proprietary
SINGLE_APP
EASY
CROSS_PLATFORM
no
2025-07-11 05:02:15.259
["Open the pre designed template and list every pre-designed page (here Shop,Services,Gallery,Contac(...TRUNCATED)
"{\"id\": 264800830346283, \"node\": \"DESKTOP-GFL3B1N\", \"model\": \"20FXS1BP00\", \"system\": \"W(...TRUNCATED)
[{"src":"https://datasets-server.huggingface.co/assets/anaisleila/computer-use-data-psai/--/{dataset(...TRUNCATED)
videos/cmcc8u6ym018l1p1yxhf18gc2.mp4
dom_snaps/cmcc8u6ym018l1p1yxhf18gc2.zip
"[{\"time_stamp\": 1209196.8302786, \"action\": \"move\", \"x\": 1459, \"y\": 60}, {\"time_stamp\": (...TRUNCATED)
cmcc8u6yd00wv1p1yy8guorre
cmcc8u68d006m1p1y4wroiffs
"In the Lighting department, filter for Pendant Lights that are β€œBlack” in finish and β€œIndustr(...TRUNCATED)
BROWSER_TASK
[ "Shopping & E-commerce" ]
Home Depot
["Navigation & Workflow Control:Navigation","Information Retrieval & Analysis:Filtering","Informatio(...TRUNCATED)
Proprietary
SINGLE_APP
EASY
CROSS_PLATFORM
no
2025-07-10 20:32:54.783
[]
"{\"id\": 264800830346283, \"node\": \"DESKTOP-GFL3B1N\", \"model\": \"20FXS1BP00\", \"system\": \"W(...TRUNCATED)
[{"src":"https://datasets-server.huggingface.co/assets/anaisleila/computer-use-data-psai/--/{dataset(...TRUNCATED)
videos/cmcc8u6yd00wv1p1yy8guorre.mp4
dom_snaps/cmcc8u6yd00wv1p1yy8guorre.zip
"[{\"time_stamp\": 1178644.7487036, \"action\": \"move\", \"x\": 1480, \"y\": 57}, {\"time_stamp\": (...TRUNCATED)
cmcc8u6yd00wr1p1yj7aot3ae
cmcc8u68c006i1p1yghwov4zh
"Use the Store Finder, enter ZIP code 10001, and note the distance (in miles) to the nearest Home De(...TRUNCATED)
BROWSER_TASK
[ "Navigation & Maps" ]
Home Depot
["Navigation & Workflow Control:Navigation","Transactional Operations:Form Filling","Information Ret(...TRUNCATED)
Proprietary
SINGLE_APP
EASY
CROSS_PLATFORM
no
2025-07-10 20:20:47.459
[]
"{\"id\": 264800830346283, \"node\": \"DESKTOP-GFL3B1N\", \"model\": \"20FXS1BP00\", \"system\": \"W(...TRUNCATED)
[{"src":"https://datasets-server.huggingface.co/assets/anaisleila/computer-use-data-psai/--/{dataset(...TRUNCATED)
videos/cmcc8u6yd00wr1p1yj7aot3ae.mp4
dom_snaps/cmcc8u6yd00wr1p1yj7aot3ae.zip
"[{\"time_stamp\": 1177951.2447762, \"action\": \"move\", \"x\": 1468, \"y\": 47}, {\"time_stamp\": (...TRUNCATED)
cmcc8u6yd00wk1p1y8d5eo7xv
cmcc8u68c006b1p1y1mgumjzd
"Open the Ceiling Fan Buying Guide in the β€œDIY Projects & Ideas” section and note the recommende(...TRUNCATED)
BROWSER_TASK
[ "Education & Learning" ]
Home Depot
[ "Navigation & Workflow Control:Navigation", "Information Retrieval & Analysis:Extraction" ]
Proprietary
SINGLE_APP
EASY
CROSS_PLATFORM
no
2025-07-10 19:35:31.038
[]
"{\"id\": 264800830346283, \"node\": \"DESKTOP-GFL3B1N\", \"model\": \"20FXS1BP00\", \"system\": \"W(...TRUNCATED)
[{"src":"https://datasets-server.huggingface.co/assets/anaisleila/computer-use-data-psai/--/{dataset(...TRUNCATED)
videos/cmcc8u6yd00wk1p1y8d5eo7xv.mp4
dom_snaps/cmcc8u6yd00wk1p1y8d5eo7xv.zip
"[{\"time_stamp\": 1175015.3670867, \"action\": \"move\", \"x\": 1095, \"y\": 328}, {\"time_stamp\":(...TRUNCATED)
cmcc8u6yd00wj1p1ytaabydeb
cmcc8u68c006a1p1y93kiyj9o
"In the Interior Paint category, apply filters to show only BEHR-brand, satin-finish, 1-gallon conta(...TRUNCATED)
BROWSER_TASK
[ "Shopping & E-commerce" ]
Home Depot
[ "Information Retrieval & Analysis:Filtering", "Information Retrieval & Analysis:Extraction" ]
Proprietary
SINGLE_APP
EASY
CROSS_PLATFORM
no
2025-07-10 18:59:55.263
[]
"{\"id\": 264800830346283, \"node\": \"DESKTOP-GFL3B1N\", \"model\": \"20FXS1BP00\", \"system\": \"W(...TRUNCATED)
[{"src":"https://datasets-server.huggingface.co/assets/anaisleila/computer-use-data-psai/--/{dataset(...TRUNCATED)
videos/cmcc8u6yd00wj1p1ytaabydeb.mp4
dom_snaps/cmcc8u6yd00wj1p1ytaabydeb.zip
"[{\"time_stamp\": 1173014.6018928, \"action\": \"move\", \"x\": 1476, \"y\": 55}, {\"time_stamp\": (...TRUNCATED)
End of preview. Expand in Data Studio

Computer Use Dataset - PSAI

A large-scale, multimodal dataset of human-computer interactions for training and evaluating AI agents.

Dataset on HuggingFace License: MIT

πŸ”— Access Dataset: https://huggingface.co/datasets/anaisleila/computer-use-data-psai


πŸ“Š Dataset Overview

This dataset contains 3,167 completed tasks of human-computer interactions captured with video, screenshots, DOM snapshots, and detailed interaction events. Created by Paradigm Shift AI for advancing computer use AI agent research.

Key Statistics

Scale:

  • 3,167 tasks with multimodal data
  • 7.87 GB dataset parquet (includes embedded screenshots)
  • 49.2 GB total (7.87 GB parquet + 16.9 GB videos + 24.4 GB DOM snapshots)
  • 100% video coverage (all 3,167 tasks)

Task Distribution:

  • Browser Tasks: 2,220 (70.1%)
  • Computer Tasks: 947 (29.9%)
  • Difficulty: Easy (79.4%) | Medium (16.7%) | Hard (3.9%)
  • Platforms: Cross-platform (95.1%) | Windows (4.5%) | macOS (0.4%)

Data Coverage by Modality

Videos: 100% coverage (3,167/3,167 tasks) - 16.9 GB
All tasks have screen recordings in MP4 format.

Screenshots: 42.6% coverage (1,349/3,167 tasks)
14,740 images embedded directly in the parquet files (included in the 7.87 GB dataset size).

DOM Snapshots: 55.8% coverage (1,766/3,167 tasks) - 24.4 GB
HTML structure captures for web-based tasks.

  • Browser tasks: 77.5% have DOM snapshots
  • Computer tasks: 4.8% have DOM snapshots

Content Diversity

  • 294 unique websites (browser tasks) - Amazon, Google, ArXiv, Apple, Booking, and more
  • 173 unique applications (computer tasks) - MS Office Suite, File Explorer, Email clients, and more
  • 31 subcategories spanning:
    • Search & Research (928 | 29.3%)
    • Shopping & E-commerce (490 | 15.5%)
    • Social Media & Communication (210 | 6.6%)
    • News & Media (149 | 4.7%)
    • Document Editing (127 | 4.0%)
    • Education & Learning (101 | 3.2%)
    • Navigation & Maps (93 | 2.9%)
    • Email Ops (71 | 2.2%)
    • And 23 more categories...

πŸš€ Quick Start

Option 1: Load Dataset Only (7.87 GB)

Fast access to metadata and embedded screenshots:

from datasets import load_dataset

# Load the dataset
dataset = load_dataset("anaisleila/computer-use-data-psai")

# Access a task
task = dataset['train'][0]
print(f"Task: {task['task_name']}")
print(f"Category: {task['category']}")
print(f"Screenshots: {len(task['screenshots'])} images")

Option 2: Download Videos & DOMs On-Demand

Download specific files as needed:

from huggingface_hub import hf_hub_download

# Download a specific video
video_path = hf_hub_download(
    repo_id="anaisleila/computer-use-data-psai",
    filename=task['video_file'],  # e.g., "videos/{task_id}.mp4"
    repo_type="dataset"
)

# Download DOM snapshot
dom_path = hf_hub_download(
    repo_id="anaisleila/computer-use-data-psai",
    filename=task['dom_snaps_file'],  # e.g., "dom_snaps/{task_id}.zip"
    repo_type="dataset"
)

Option 3: Clone Full Dataset (49.2 GB)

Clone everything including videos and DOM files:

git lfs install
git clone https://huggingface.co/datasets/anaisleila/computer-use-data-psai

πŸ“š What's in Each Task

Each task includes:

Metadata Fields

  • unique_data_id (string): Unique identifier for each recording
  • taskId (string): Task template ID (non-unique - same task done by different vendors)
  • task_name (string): Human-readable task description
  • category (string): BROWSER_TASK or COMPUTER_TASK
  • subCategory (list[string]): Specific categories (e.g., "Search & Research")
  • application_website (string): Application or website used
  • tags (list[string]): Descriptive tags
  • benchmark (string): Benchmark identifier
  • appType (string): SINGLE_APP or MULTI_APP
  • difficulty (string): EASY, MEDIUM, or HARD
  • os (string): CROSS_PLATFORM, WINDOWS, macOS, or LINUX
  • requires_login (string): Whether task requires authentication
  • completedAt (string): Timestamp (ISO 8601 format)

Multimodal Data

  • screenshots (list[images]): Screenshots at key moments - embedded and viewable
  • video_file (string): Path to screen recording (MP4) - download on demand
  • dom_snaps_file (string): Path to HTML DOM snapshot (ZIP) - download on demand
  • events (string): Keyboard/mouse interactions with timestamps (JSON)
  • reasoning_steps (list[string]): Step-by-step task completion reasoning
  • metadata (string): System info (OS, screen resolution, hardware) (JSON)

Note: Screenshots are embedded for instant browsing. Videos and DOM snapshots are stored separately to keep the dataset size manageable.


πŸ’‘ Usage Examples

1. Browse and Explore Tasks

from datasets import load_dataset
import json

dataset = load_dataset("anaisleila/computer-use-data-psai")

# Browse tasks
for task in dataset['train'][:5]:
    print(f"Task: {task['task_name']}")
    print(f"  Category: {task['category']}")
    print(f"  Difficulty: {task['difficulty']}")
    
    # Parse metadata
    metadata = json.loads(task['metadata'])
    print(f"  System: {metadata.get('system')}")
    
    # Parse events
    if task['events']:
        events = json.loads(task['events'])
        print(f"  Events: {len(events)} interactions")

2. Filter by Criteria

# Filter by difficulty
hard_tasks = dataset['train'].filter(lambda x: x['difficulty'] == 'HARD')
print(f"Hard tasks: {len(hard_tasks)}")

# Filter by category
browser_tasks = dataset['train'].filter(lambda x: x['category'] == 'BROWSER_TASK')

# Complex filter
windows_hard = dataset['train'].filter(
    lambda x: x['difficulty'] == 'HARD' and x['os'] == 'WINDOWS'
)

3. Download Files for Specific Tasks

from huggingface_hub import hf_hub_download

# Find a task you're interested in
task = dataset['train'][0]

# Download video
video = hf_hub_download(
    repo_id="anaisleila/computer-use-data-psai",
    filename=task['video_file'],
    repo_type="dataset"
)

# Download DOM snapshot (if available)
if task['dom_snaps_file']:
    dom = hf_hub_download(
        repo_id="anaisleila/computer-use-data-psai",
        filename=task['dom_snaps_file'],
        repo_type="dataset"
    )

🎯 Use Cases

This dataset supports:

  • Training computer use AI agents (vision-language-action models)
  • Reinforcement learning for GUI interaction
  • Benchmark evaluation of computer use capabilities
  • Research in human-computer interaction patterns
  • Accessibility tools development
  • Software testing and quality assurance automation

πŸ“– Dataset Details

Data Collection

Data was collected using a custom-built computer interaction capture tool that records:

  • Keyboard and mouse inputs with timestamps
  • Full screen video recordings
  • DOM snapshots for web-based tasks
  • Accessibility tree information
  • Detailed event streams

Human vendors performed tasks following specific instructions. All vendors signed disclosure agreements authorizing public release of the data.

Privacy & Consent

  • Data collected from consenting human vendors
  • Vendors signed disclosure agreements for public release
  • May contain some PII from vendor interactions
  • Users should be aware tasks may show personal information

Known Limitations

  • Some tasks may reference applications or websites that have changed since data collection
  • Not all tasks have screenshots or DOM snapshots (see coverage stats above for exact percentages)
  • Dataset contains 100 duplicate rows (3,267 total rows, 3,167 unique tasks)
    • To deduplicate: dataset.to_pandas().drop_duplicates(subset=['unique_data_id'], keep='first')

πŸ“œ License

MIT License - see LICENSE for full details.

Copyright (c) 2025 Paradigm Shift AI
Anais Howland, Ashwin Thinnappan, Jameel Shahid Mohammed

πŸ™ Citation

If you use this dataset in your research, please cite:

@dataset{psai_computer_use_2025,
  title={Computer Use Data - Paradigm Shift AI},
  author={Anais Howland and Ashwin Thinnappan and Jameel Shahid Mohammed},
  organization={Paradigm Shift AI},
  year={2025},
  publisher={HuggingFace},
  url={https://huggingface.co/datasets/anaisleila/computer-use-data-psai}
}

πŸ‘₯ Authors

Anais Howland, Ashwin Thinnappan, and Jameel Shahid Mohammed
Paradigm Shift AI

This dataset was created by the team at Paradigm Shift AI, including:

  • Data collection infrastructure and vendor coordination system
  • Custom screen recording and interaction capture tool
  • Dataset curation, validation, and quality assurance

πŸ“ž Contact & Contributions

This dataset is provided as-is for the research community.

For questions or issues:


πŸ“‹ Changelog

  • v1.0 (2025): Initial public release with 3,167 tasks
Downloads last month
1,421