AReaL (Ant Reasoning RL) is an open-source fully asynchronous reinforcement learning training system for large reasoning models developed at the RL Lab, Ant Research. Built upon the open-source project RealHF, we are fully committed to open-source by providing training details, data, and infrastructure required to reproduce results along with the model itself. AReaL aims to help everyone build their own AI agents easily and affordably. Our team loves milk tea because it's delicious, customizable, and affordable. We hope you enjoy our project just like how you enjoy real-world milk tea (cheers).
AReaL Highlights
🔥 [NEW] Asynchronous RL: With algorithm-system co-design, AReaL supports fully asynchronous RL for the fastest training! Experimental support for multi-turn agentic RL is also provided.
🛠️ Open & Reproducible: We continuously release all code, datasets, and training recipes for RL training of LLMs.
🚀 Scalability: AReaL can seamlessly adapt to different computational resource settings, ranging from a single node to 1K GPUs.
🔪 Cutting-Edge Performance: AReaL can produce models with cutting-edge reasoning capabilities in math and coding. We are also actively working on agentic tasks.
News
[2025/06/03] (v0.3, boba²) We release boba² (double-boba) for fully asynchronous RL training, which achieves a 2.77x speedup while obtaining on-par or even better training performance compared to synchronous systems. Moreover, asynchronous RL makes it extremely easy to set up multi-turn agentic RL training! Check out our v0.3 overview blog and the research paper.
[2025/03/31] (v0.2, Boba) Here comes our next milestone release - Boba! Please call it A-ReaL-boba! This release includes much faster training with SGLang support and SOTA 7B and 32B models on math reasoning. Check our v0.2 technical blog.
[2025/02/24] (v0.1) Our initial release includes reproducible results for 1.5B and 7B LRMs. Check our v0.1 technical blog.
Release Highlights
In our AReaL-boba² (A-ReaL-double-boba) release, we highlight the top 3 most important features:
A fully asynchronous RL training pipeline with system and RL algorithm co-design, achieving over 2.77x speedup without any performance drop. Check the benchmark scripts and instructions here.
SOTA coding models, i.e., a 14B model with a 69.1 score on LCB-v5. To reproduce, check the configs and instructions.
Experimental support for multi-turn agentic RL training. Check our complete example.
For the complete system design and more training details, please check our v0.3 blog and our research paper.
Overview of Asynchronous RL Training
During the synchronous RL training process, a generation step must wait until the longest sequence completes within the batch of LLM outputs. Due to the varying output lengths for LRMs, a synchronous RL system suffers from massive GPU idle time, leading to training inefficiency. Some recent works (DeepCoder, Intellect) propose overlapping a single training step with a single generation step to accelerate training. However, the largest bottleneck remains unchanged: the samples within a batch are still from the same model version, leading to waiting and GPU idle time.
Synchronous vs One-step Overlap RL
Fig.1. Left: Execution timeline of synchronous RL training. Right: Execution timeline of one-step overlap RL system.
AReaL adopts a fully asynchronous RL training framework that completely decouples generation from training. In AReaL, LLM generation runs in a streaming manner, with each rollout worker continuously producing outputs without waiting. Meanwhile, trainer workers perform parallel model updates upon receiving training batches.
Asynchronous RL Training
Fig 2. Execution timeline of our fully asynchronous RL system.
AReaL follows a system-algorithm co-design principle: on the system side, AReaL efficiently syncs model parameters and carefully controls the staleness of each training sample; on the algorithm side, AReaL improves the objective of PPO to make async-RL stable.
We compare the scalability of asynchronous RL training based on our AReaL-boba² system with classical synchronous RL training (we adopt the fastest open-source system veRL, main branch on 05/07/2025) across different model sizes and different numbers of H800 GPUs. AReaL demonstrates much improved scaling capabilities with respect to training throughput. This is also partially due to AReaL decoupling training and generation, leading to much fewer GPU memory fragments.
Scaling Comparison
Fig.3 The scaling trend of asynchronous RL (based on AReaL-boba2) and classical synchronous RL (based on veRL) with different model sizes. Dotted lines indicate ideal linear scaling.
SOTA Code Generation Model by AReaL-boba²
We use Qwen3 as our base model. After asynchronous RL training, we achieve SOTA results on LiveCodeBench, Codeforces, and CodeContests benchmarks.
Table 1: Coding Task Performance Comparison. AReaL-boba²-8B/14B-Open denotes training results on open-source data. AReaL-boba²-8B/14B models are trained with an additional small amount of internal data and achieve SOTA performance on LiveCodeBench, Codeforces & CodeContests.
We highlight the tutorials and code walkthroughs about the following key features for asynchronous training:
In particular, we show a simple example to develop a multi-turn math agent for RL training. Please see the learning curve below and reference the step-by-step guide if you want to implement your own agentic RL project.
AReaL is under active development. We plan to have minor releases weekly and major releases monthly. Community engagement and contributions are extremely welcome. We are also hiring interns and full-time employees with open positions in both the US and China.
For the research and development plan already in place, please see the following list:
System Development
Support for SGLang
RL training with coding problems
Asynchronous generation and RL training
Optimizations for distributed training: expert parallel for MOE and zero-bubble pipelining
RL for vision-language models (VLM)
Multi-turn agentic RL
Function calling and tool use
Algorithm Development
RL training recipes for 1.5B and 7B models
A complete RL training recipe for 32B models
Sample-efficient multi-task RL algorithms
Agentic capabilities with end-to-end RL
Stable RL training for larger MOE models
Acknowledgement
We would like to note that major contributors are from the RL Lab at Ant Research and the Institute for Interdisciplinary Information Sciences, Tsinghua University.
Our team has also received invaluable assistance from the Data Intelligence Lab at Ant Research for data support and from the Super Computing Technology (SCT) team at Ant Group, particularly in the realm of large-scale cluster operations and maintenance.
1@inproceedings{mei2025real,
2 author = {Mei, Zhiyu and Fu, Wei and Li, Kaiwei and Wang, Guangju and Zhang, Huanchen and Wu, Yi},
3 title = {ReaL: Efficient RLHF Training of Large Language Models with Parameter Reallocation},
4 booktitle = {Proceedings of the Eighth Conference on Machine Learning and Systems,
5 MLSys 2025, Santa Clara, CA, USA, May 12-15, 2025},
6 publisher = {mlsys.org},
7 year = {2025},
8}
bibtex
1@misc{fu2025areal,
2 title={AReaL: A Large-Scale Asynchronous Reinforcement Learning System for Language Reasoning},
3 author={Wei Fu and Jiaxuan Gao and Xujie Shen and Chen Zhu and Zhiyu Mei and Chuyi He and Shusheng Xu and Guo Wei and Jun Mei and Jiashu Wang and Tongkai Yang and Binhang Yuan and Yi Wu},
4 year={2025},
5 eprint={2505.24298},
6 archivePrefix={arXiv},
7 primaryClass={cs.LG},
8 url={https://arxiv.org/abs/2505.24298},
9}
