AgentIndex icon
AgentIndex
ToolsCategoriesTrendingNewCompare
Submit Tool
Home/
LLM Infra/
AReaL
AReaL logo

AReaL

Active·★ 5.2k·Updated 2026-05-29
★ Trending

Lightning-Fast RL for LLM Reasoning and Agents. Made Simple & Flexible.

AReaL is an open-source, fully asynchronous reinforcement learning training system designed for large reasoning and agentic models. It offers exceptional flexibility, industry-leading speed, and scalability from a single node to over 1,000 GPUs, achieving state-of-the-art performance.

#Reinforcement Learning#Large Language Models#Asynchronous Systems#Agentic AI#Scalable Training
$ Install
$ pip install -e ".[dev,docs]"
↗ Visit site★ GitHub
01

Features

01Fully Asynchronous RL Training: Enables stable, industry-leading speed for reinforcement learning.
02Scalability: Seamlessly adapts from single-node setups to over 1,000 GPUs.
03Flexible Agentic Rollout: Easy customization for multi-turn agentic workflows and integration with external frameworks.
04Cutting-Edge Performance: Achieves state-of-the-art results for math, coding, and search agents.
05Open-Source & Reproducible: Provides full training details, data, and infrastructure to reproduce results.
02

Compatibility

Ascend NPU
Stable Support
Verified via docs
Megatron
Partial Integration
Verified via docs
PyTorch FSDP
Partial Integration
Verified via docs
vLLM
Partial Integration
Verified via docs
SGLang
Partial Integration
Verified via docs
03

Quick start

1
$ pip install -e ".[dev,docs]"
04

Use cases

↳Training Reasoning Agents: Developing AI agents capable of complex mathematical, coding, and general reasoning tasks.
↳Large Language Model Alignment (RLHF): Fine-tuning LLMs using Reinforcement Learning from Human Feedback.
↳Multi-Turn Agentic Workflows: Implementing and customizing iterative agent behaviors with self-correction and tool integration.
05

Alternatives

MetaGPT logo
MetaGPT★ 68.4k
🌟 The Multi-Agent Framework: First AI Software Company, Towards Natural Language Programming
vs →
cua logo
cua★ 17.3k
Open-source infrastructure for Computer-Use Agents. Sandboxes, SDKs, and benchmarks to train and evaluate AI agents that can control full desktops (macOS, Linux, Windows).
vs →
agentscope logo
agentscope★ 25.8k
AgentScope: Agent-Oriented Programming for Building LLM Applications
vs →
ChatGPT on WeChat logo
ChatGPT on WeChat★ 44.9k
Empower your WeChat with ChatGPT. Supports text, voice, and image generation.
vs →
devin.cursorrules logo
devin.cursorrules★ 6.0k
Magic to turn Cursor/Windsurf as 90% of Devin
vs →
rllm logo
rllm★ 5.6k
Democratizing Reinforcement Learning for LLMs
vs →
ir-sim logo
ir-sim★ 1.1k
A Python-based lightweight robot simulator designed for navigation, control, and reinforcement learning
vs →
langchain4j-aideepin logo
langchain4j-aideepin★ 1.3k
基于AI的工作效率提升工具(聊天、绘画、知识库、工作流、 MCP服务市场、语音输入输出、长期记忆) | Ai-based productivity tools (Chat,Draw,RAG,Workflow,MCP marketplace, ASR,TTS, Long-term memory etc)
vs →
See all alternatives →

Related searches

AReaL AlternativesBest LLM Infra Tools 2026Open Source LLM InfraAReaL TutorialAReaL Vs CompetitorsReinforcement LearningLarge Language ModelsAsynchronous Systems

Comments

Log in to leave a comment

No comments yet. Be the first!

On this page
01Features02Compatibility03Quick start04Use cases05Alternatives
Stats
GitHub Stars★ 5.2k
Last commit1d ago
StatusActive
License—
CategoryLLM Infra
Trend (30d)
+0.2k↑ 4.1%
Links
Documentation↗Discussion↗Issues↗Releases↗

Deploy on DigitalOcean — Get $200 Free Credit

Ad
© 2026 AgentIndex.app|Built by a 10-year iOS Developer.
QYSGitHubBuy me a coffee ☕

Browse by Category

Code AssistantWorkflow AutomationRAG / Knowledge BaseMulti-AgentBrowser AutomationLLM InfraDev ToolingObservability

Not affiliated with Anthropic, OpenAI or Microsoft.