OpenEvolve is an open-source implementation of Google DeepMind's AlphaEvolve system, designed for researchers and developers seeking to automate code optimization and discovery. It leverages Large Language Models (LLMs) in an evolutionary process to iteratively improve code across multiple programming languages, aiming to achieve state-of-the-art results in complex problem-solving.

How It Works

OpenEvolve employs an evolutionary algorithm orchestrated by an asynchronous pipeline. Key components include a Prompt Sampler for generating context-rich prompts, an LLM Ensemble for code generation, an Evaluator Pool for testing and scoring programs, and a Program Database for storing evolution history. This architecture allows for the efficient exploration of the solution space by generating, evaluating, and selecting code modifications, enabling multi-objective optimization and flexible prompt engineering.

Quick Start & Requirements

• Installation: pip install -e . after cloning the repository.
• Prerequisites: Python, OpenAI-compatible LLM API access.
• Usage: Python API (OpenEvolve class) or command-line interface (python openevolve-run.py).
• Resources: Docker image available.
• Examples & Docs: Available in the examples/ directory and via the project's GitHub repository.

Highlighted Details

• Evolves entire code files, not just functions.
• Supports multiple programming languages and OpenAI-compatible LLM APIs.
• Achieved state-of-the-art results in circle packing problems.
• Demonstrates transformation of simple algorithms into sophisticated ones (e.g., random search to simulated annealing).

Maintenance & Community

• Primarily developed by Asankhaya Sharma.
• Project hosted on GitHub.

Licensing & Compatibility

• License: MIT.
• Compatibility: Permissive for commercial use and integration with closed-source projects.

Limitations & Caveats

The project is described as an implementation of a 2025 paper, suggesting it may represent cutting-edge research but could also be subject to rapid development and potential instability. Specific LLM model compatibility and performance tuning may require significant configuration effort.