Privacy and advertising choices

Git-Stars uses essential storage for site operation. Optional analytics and ad-measurement scripts stay disabled unless you accept them; partners such as Google may then use cookies or similar identifiers where required. Privacy Policy

LogoGit-Stars
Top StarsTrendingAI AgentsDaily PicksViral ReposInsights
LogoGit-Stars

Discover top GitHub projects with real rankings and AI insights

GitHub
Built withLogo of Git-StarsGit-Stars
Rankings
  • Top Stars
  • Trending
  • AI Agents
  • Daily Picks
  • Explore
Resources
  • Insights
  • Editorial Policy
About
  • About
  • Contact
Legal
  • Privacy Policy
  • Terms of Service
© 2026 Git-Stars. All Rights Reserved.
AI Agent Analysis
PR

PrimeIntellect-ai/prime-rl

Agentic RL Training at Scale

stars
1.9k
Language
Python
GitHub
Source and compliance noteLast synced: Aug 15, 2026

Git-Stars is independent and not affiliated with GitHub or this project. Analysis may be AI-assisted and based on public repository metadata plus short README-derived summaries. We do not mirror full README files, docs, issues, or social comments.

Original GitHub sourceMethodologyEditorial Policy

Overview

prime-rl is a framework for large-scale reinforcement learning (RL) training of AI agents, designed to be easy to use and hackable while scaling to 1000+ GPUs. It emphasizes fully asynchronous RL for high-throughput agentic training, with native integration with verifiers environments and support for multimodal models.

Installation

curl -sSL https://raw.githubusercontent.com/PrimeIntellect-ai/prime-rl/main/scripts/install.sh | bash

Problem solved

It addresses the challenge of training large-scale agentic RL models efficiently by providing a fully asynchronous, high-throughput framework that scales to 1000+ GPUs. It solves the complexity of managing distributed training with FSDP2, vLLM inference, and multi-node deployment, while supporting a wide range of model families including MoE and multimodal models.

What you can build

Developers can build and train large-scale RL agents for tasks like software engineering (SWE), agentic environments, and multimodal reasoning. The framework supports end-to-end post-training (SFT, RL, evals), and can handle frontier models like GLM-5 with advanced features like FP8 inference, PD disaggregation, and Mooncake KV offload. The ceiling includes training 1T+ parameter MoE models on 1000+ GPUs with expert and context parallelism.

Community sentiment

Positive

No community feedback yet.

Concerns

No concerns documented yet.

Bottom line

This framework is ideal for researchers and organizations that need to train large-scale RL agents at scale, especially those working with MoE or multimodal models. It is not suitable for those without access to NVIDIA GPUs or who need a simple, low-resource setup. The key trade-off is its complexity and hardware requirements in exchange for extreme scalability and performance.

Analyzed by Git-Stars - 8/15/2026