#107107th of 118 in Agent frameworks and SDKs
rlm
Python library that lets an LLM recursively call itself from a REPL
Documentation ↗
(opens in a new tab)Website ↗
(opens in a new tab)Repository on GitHub ↗
(opens in a new tab)
- Stars
- 5.8k
- License
- MIT
- Last commit
- Nov 2026
- Last release
- Jun 2026
- Language
- Python
3,469 on 21 Apr 20265,759 on 10 Nov 2026
From GitHub's daily star counts.
Checked on GitHub.
Overview
rlms replaces a plain llm.completion call with rlm.completion. The prompt is loaded as a variable in a REPL, and the model writes code to inspect it and launch sub-LM calls. It supports OpenAI, Anthropic, OpenRouter and Portkey clients, plus local models through vLLM's OpenAI interface. The repo also includes a verifiers/prime-rl training environment and a trajectory visualizer.
Who it is for: Developers and researchers building long-context LLM agents in Python
Strengths
- Handles long contexts by keeping the prompt as a REPL variable with recursive sub-calls
- REPL backends: local, ipython, docker, modal, prime, daytona, e2b
- Includes a prime-rl/verifiers training setup with an OOLONG example
- JSONL trajectory logging with a local visualizer for inspecting sub-calls
Weaknesses
- Default local REPL runs Python exec in the host process; README says not for production
- Prime Sandboxes are beta, with slow runtimes noted as an open issue
- Cloud sandbox environments need separate accounts and API keys (Modal, Prime, Daytona, E2B)
- Training harness uses unsandboxed local REPL, which the README says is not ideal for safety
What it needs
- no GPU
- Needs Python 3.11+, LLM API or vLLM server, Docker (optional), Modal/Prime/Daytona/E2B (optional), Node.js (visualizer)
- Models: OpenAI, Anthropic, OpenRouter, Portkey, vLLM
- port 3001
Quick answers
Is rlm open source?
Yes. Its code is on GitHub under the MIT license.
Does rlm need a GPU?
No, it runs without one.
Also in Agent frameworks and SDKs
See all 118| Rank | Project | Score |
|---|---|---|
| 1 | ppt-masterAgent skill that turns documents into natively editable PowerPoint decks | 95 out of 100 |
| 2 | openai-agents-pythonPython SDK for multi-agent workflows with handoffs, guardrails and tracing | 91 out of 100 |
| 3 | mastraTypeScript framework for building AI agents, workflows and MCP servers | 91 out of 100 |
| #4 | browser-usePython library that lets an LLM agent control a web browser | 90 out of 100 |
| #5 | copilotkitSDK for adding agent chat, generative UI and shared state to apps | 90 out of 100 |