#2222nd of 28 in Agent platforms
rd-agent
LLM agent that proposes and implements ML experiments for quant and data science
Live demo ↗
(opens in a new tab)Documentation ↗
(opens in a new tab)Repository on GitHub ↗
(opens in a new tab)
- Stars
- 14.9k
- License
- MIT
- Last commit
- Sep 2026
- Last release
- Sep 2026
- Language
- Python
Missed the latest GitHub check; these facts may be older.
Overview
RD-Agent splits automated R&D into a research agent that proposes ideas and a development agent that implements and tests them. Scenarios include Qlib-based quant factor and model evolution, Kaggle and MLE-bench data science runs, paper-to-model extraction, and LLM fine-tuning. LLM access goes through LiteLLM, and it runs on Linux with Docker.
Who it is for: ML engineers and quant researchers automating experiment loops with LLMs
Strengths
- Reports MLE-bench results: 30.22% overall with o3 research and GPT-4.1 development agents
- Ready-made CLI scenarios: fin_quant, fin_factor, fin_model, general_model, data_science
- LiteLLM backend works with OpenAI, Azure OpenAI and DeepSeek, among others
- Published on PyPI with docs, a live demo and several accompanying papers
Weaknesses
- Linux only; most scenarios need Docker usable without sudo
- Needs a chat model with json_mode plus an embedding model, so API costs apply
- Data science setup requires manual dataset download and several .env variables
- Web UI excludes the data_science scenario
What it needs
- Needs Docker, Conda, Qlib, LiteLLM, LLM API with embeddings
- Models: OpenAI, Azure OpenAI, DeepSeek, LiteLLM providers, SiliconFlow embeddings
Also in Agent platforms
See all 28| Rank | Project | Score |
|---|---|---|
| 1 | LangflowVisual builder for AI agents and workflows, exposed as API or MCP server | 81 out of 100 |
| 2 | daily_stock_analysisDaily LLM stock analysis for A-share, HK, US and other markets, pushed to chat apps | 80 out of 100 |
| 3 | mindshubAgent workspace for knowledge work, with swappable open agent harnesses | 78 out of 100 |
| #4 | deer-flowAgent harness that runs sub-agents, memory and sandboxed skills for long tasks | 76 out of 100 |
| #5 | DifyVisual builder for LLM workflows, RAG pipelines and agents, with APIs | 75 out of 100 |