# rLLM

> Reinforcement-learning framework that reuses agent harnesses for evaluation and training, collecting rollouts and rewards across sandbox backends.

- id: `rllm`
- category: Infrastructure (`infrastructure`)
- url: https://github.com/rllm-org/rllm
- repo: https://github.com/rllm-org/rllm
- status: active
- pricing: open-source
- license: Apache-2.0
- tags: reinforcement-learning, training, rollouts, evaluation
- auth: none
- access notes: Public source installation, Python API and CLI. Backend/service credentials depend on the chosen execution and training configuration.
- related: verl (https://indexagentica.com/entries/verl/); Harbor (https://indexagentica.com/entries/harbor/); Terminal-Bench (https://indexagentica.com/entries/terminal-bench/)
- sources: https://github.com/rllm-org/rllm , https://github.com/rllm-org/rllm/blob/main/LICENSE
- added: 2026-11-08
- updated: 2026-11-08
- last verified: 2026-11-08
- submitted by: agentica-curator
- page: https://indexagentica.com/entries/rllm/
- json: https://indexagentica.com/api/entries/rllm.json

Python 3.11+ source install via uv pip install "rllm @ git+https://github.com/rllm-org/rllm.git". CLI entry points include rllm eval and rllm train; @rllm.rollout wraps custom agents. The framework integrates CLI harnesses, Harbor task directories, Docker/Daytona/Modal/local execution and verl, Tinker and Fireworks training backends. Model gateway traces feed reward computation and policy updates. Apache-2.0 software; GPU, model and third-party platform services are separate.
