Overview
News
Technologies
Salaries
Products
People
Growth
Financials
Overview
Training the next generation of intelligent agents.. rLLM is an open-source initiative building the infrastructure to train, evaluate, and evolve intelligent agents at scale.
News
Continual Learning via Real-Time RL for Agents
We study single-rollout learning in a real-time RL setup and show that a simple batch-normalized advantage lifts Qwen3-Coder-30B on MigrationBench from 43% to 59.2% - a 16.2% absolute gain - from one rollout per task.
Read more
Report
Training a Frontier Java Code Migration Agent with AWS AgentCore Runtime
We train Qwen3-Coder-30B-A3B using AgentCore Runtime integration to outperform Claude 4.5 Haiku on long-horizon Java repo migrations.
Read more
Report
rLLM UI: Real-Time Observability for Agent Training & Evaluation
A real-time observability platform for training and evaluating agents. Other tools show what is happening during training - rLLM UI shows you why, letting you inspect exactly what the model generates at every step.
Read more
Report
On-Policy Distillation: Training Smaller Students from Stronger Teachers
rLLM On-Policy Distillation (OPD) trains smaller students from stronger teachers by using the teacher's policy to guide the student's training - a practical recipe for compact, capable models.
Read more
Report
Pro access
Upgrade to see all 11 mentions
Upgrade to a paid plan to read every media mention of this company - funding news, awards, product launches and press releases from all the outlets writing about it.
Every media mention and press release
Funding news, awards and product launches
Fresh coverage from every outlet writing about the company
Upgrade now
Cancel anytime. Secure checkout. Instant activation.