---
license: cc-by-4.0
task_categories:
- other
dataset_info:
- config_name: all_pull_request
features:
- name: id
dtype: int64
- name: number
dtype: int64
- name: title
dtype: string
- name: user
dtype: string
- name: user_id
dtype: int64
- name: state
dtype: string
- name: created_at
dtype: string
- name: closed_at
dtype: string
- name: merged_at
dtype: string
- name: repo_url
dtype: string
- name: repo_id
dtype: int64
- name: html_url
dtype: string
- name: body
dtype: string
- name: agent
dtype: string
configs:
- config_name: all_pull_request
data_files:
- split: train
path: all_pull_request.parquet
- config_name: all_repository
data_files:
- split: train
path: all_repository.parquet
- config_name: all_user
data_files:
- split: train
path: all_user.parquet
- config_name: pull_request
data_files:
- split: train
path: pull_request.parquet
- config_name: repository
data_files:
- split: train
path: repository.parquet
- config_name: pr_comments
data_files:
- split: train
path: pr_comments.parquet
- config_name: pr_reviews
data_files:
- split: train
path: pr_reviews.parquet
- config_name: pr_review_comments
data_files:
- split: train
path: pr_review_comments.parquet
- config_name: pr_commits
data_files:
- split: train
path: pr_commits.parquet
- config_name: pr_commit_details
data_files:
- split: train
path: pr_commit_details.parquet
- config_name: pr_timeline
data_files:
- split: train
path: pr_timeline.parquet
- config_name: issue
data_files:
- split: train
path: issue.parquet
- config_name: related_issue
data_files:
- split: train
path: related_issue.parquet
---
# AIDev-7.6M — snapshot of AIDev v5
> **This repository is a frozen snapshot of AIDev v5 (cutoff date of March 31, 2026).**
> The dataset is maintained in the main repository, **https://huggingface.co/datasets/hao-li/AIDev**,
> where the full dataset card lives (call for papers, papers using AIDev, example notebooks,
> memory-safe loading guide, mining queries, update history). The data here is byte-identical to
> the [`v5` tag](https://huggingface.co/datasets/hao-li/AIDev/tree/v5) of the main repository.
> **Please use and cite the main repository; this one will not be updated.**
>
> - **Main repository (latest):** https://huggingface.co/datasets/hao-li/AIDev
> - **This exact version in the main repository:** https://huggingface.co/datasets/hao-li/AIDev/tree/v5
> - **Papers:** [The Rise of AI Teammates in SE 3.0](https://huggingface.co/papers/2507.15003) · [AIDev: Studying AI Coding Agents on GitHub](https://huggingface.co/papers/2602.09185)
> - **GitHub:** https://github.com/SAILResearch/AI_Teammates_in_SE3
## Loading This Version
Load from the main repository, pinned to the `v5` tag (same files as this repository):
```py
import pandas as pd
all_repo_df = pd.read_parquet("hf://datasets/hao-li/AIDev@v5/all_repository.parquet")
all_user_df = pd.read_parquet("hf://datasets/hao-li/AIDev@v5/all_user.parquet")
```
> ⚠️ `all_pull_request.parquet` is ~6.7 GB on disk and `pr_commit_details.parquet` is ~12.4 GB —
> **do not** load them blindly with a plain `pd.read_parquet`. See the
> [Loading the Dataset](https://huggingface.co/datasets/hao-li/AIDev#loading-the-dataset-without-blowing-up-your-ram)
> section of the main card and the memory-safe [`scripts/`](https://huggingface.co/datasets/hao-li/AIDev/tree/v5/scripts)
> (a copy ships in this repository as well).
## Quick Look (v5)
| Agent | # PR | % Open PR | % Closed PR | % Merged PR | # Developers | # Repository |
| --- | --- | --- | --- | --- | --- | --- |
| OpenAI Codex | 3,844,602 | 7.3 | 6.5 | 86.2 | 237,223 | 355,920 |
| Copilot | 1,581,924 | 14.2 | 15.8 | 69.9 | 0 | 347,971 |
| Claude Code | 1,567,100 | 4.2 | 7.8 | 88.0 | 114,516 | 174,038 |
| Cursor | 317,453 | 28.3 | 14.3 | 57.4 | 62,246 | 81,915 |
| Google Jules | 296,411 | 21.0 | 15.6 | 63.4 | 18 | 41,125 |
| Devin | 77,791 | 17.0 | 23.3 | 59.6 | 0 | 10,321 |
| **Total** | **7,685,281** | **9.6** | **9.5** | **80.9** | **393,349** | **961,168** |
> **Note on `# Developers`:** this counts distinct human PR authors. Copilot, Devin, and
> Google Jules PRs are authored by the agent's own bot account (e.g., `copilot`,
> `devin-ai-integration[bot]`, `google-labs-jules[bot]`), so we report 0 developers for Copilot
> and Devin by definition, and Google Jules is near zero, even though humans initiated the work.
> **Note on Google Jules volume:** `google-labs-jules[bot]` PR volume dropped sharply
> (~15×) after late January 2026 and stayed at the lower level through the March 31 cutoff.
> We verified against the live GitHub API that this is a real change in the agent's PR
> output — not missing data, and the volume did not migrate to another account.
![](AIDev_new_figure.png)
## Tables in This Version
Full scope (every repository): `all_pull_request`, `all_repository`, `all_user`.
AIDev-pop (repositories with more than 100 stars): `pull_request`, `repository`, `pr_timeline`,
`pr_comments`, `pr_reviews`, `pr_review_comments`, `pr_commits`, `pr_commit_details`, `issue`,
`related_issue`. Column-level descriptions, join keys and the agent mining queries are on the
[main dataset card](https://huggingface.co/datasets/hao-li/AIDev#dataset-structure).
Version notes specific to v5:
* The per-PR sub-resource tables (`pr_comments`, `pr_reviews`, `pr_review_comments`, `pr_commits`,
`pr_timeline`) are fetched with full pagination; v4 and v3 truncate them at 30 items per PR.
Re-fetched threads are a fresh snapshot, so filter on `created_at` if you need strict cutoff
semantics, and file-level `pr_commit_details` may be missing for commits beyond the first 30
of very large PRs.
* The `pr_task_type` table is not shipped (since v4). It is still available for v3:
`pd.read_parquet("hf://datasets/hao-li/AIDev@v3/pr_task_type.parquet")`.
## License and Citation
Repository contents maintain their original licenses; see the
[License](https://huggingface.co/datasets/hao-li/AIDev#license) section of the main card.
If you use AIDev in your work, please cite:
```bibtex
@misc{li2025aiteammates_se3,
title={The Rise of AI Teammates in Software Engineering (SE) 3.0: How Autonomous Coding Agents Are Reshaping Software Engineering},
author={Hao Li and Haoxiang Zhang and Ahmed E. Hassan},
year={2025},
eprint={2507.15003},
archivePrefix={arXiv},
primaryClass={cs.SE},
url={https://arxiv.org/abs/2507.15003}
}
```