OpenEvolve

OpenEvolve is an open-source evolutionary coding framework that automates the discovery of optimized algorithms by turning large language models into autonomous optimizers.

OpenEvolve is a library for evolutionary coding that uses large language models to optimize algorithms and discover new solutions. It employs an island-based architecture and MAP-Elites, a quality-diversity evolution approach, to maintain diverse populations and prevent premature convergence. The framework supports Python, Rust, R, and Metal shaders, and provides automated pipelines for performance optimization, scientific computing, and algorithm design. Users can configure model behavior through custom prompts, decision hooks, and intelligent ensembles. It integrates with any OpenAI-compatible API, including direct OpenAI, Google Gemini, local models, and tools like the Claude Code CLI. The system handles feature binning and evaluation, allowing developers to evolve code, prompts, or hardware kernels with reproducible results.

OpenEvolve at a glance

Best for: Software engineer, Researcher
Runs on: API, CLI, self-hosted, open source
Pricing: No direct subscription fee; users pay based on the costs of their chosen LLM provider (e.g., OpenAI or Google Gemini) per iteration.
Released: May 2025
Category: AI coding tool
Website: github.com/algorithmicsuperintelligence/openevolve

Key features

Autonomous code optimization using LLM ensembles
Island-based architecture for population management
Quality-diversity evolution via MAP-Elites
Deterministic execution for scientific reproducibility
Configurable decision hooks and selection policies
Support for OpenAI-compatible LLM backends
Automatic hardware-specific kernel optimization

Use it for

Discovering hardware-specific GPU kernel optimizations
Automating manual tuning for scientific computing tasks
Generating novel adaptive sorting algorithms
Optimizing prompts to improve LLM accuracy on benchmarks
Finding Pareto-optimal solutions for multi-objective problems

Companions

Tools that articles about OpenEvolve recommend using with it, each with its job.

Context7 — Provides up-to-date documentation for AI agents and code editors.
Ollama — Run local models at zero marginal cost, with code kept on your machine.
OpenRouter — Access over 500 models and 80 providers through one unified API.
Daytona — Runs sessions in isolated sandboxes synced to git branches.
Composio — Equips AI agents with tools, managed user auth, and sandboxes across 1,500+ apps
Oh My OpenCode Slim — Lightweight multi-agent orchestration and background tasks.

Alternatives to OpenEvolve

OpenCode — Works with 75+ model providers, local models, and existing subscriptions
Opik — It offers comprehensive tracing, automated evaluations, and open-source flexibility.
OpenChamber — Run and review multi-model AI coding agents from desktop, browser, and mobile devices.
OpenResearch — Run parallel research workflows locally or on remote compute.
EvoAgentX — It automatically builds, evaluates, and evolves multi-agent workflows.
Oumi — Automates model training and evaluation to cut costs and improve accuracy.
jevgrep — Reduces coding agent costs by providing targeted, relevant repository context.
Reactive Optimizers — It automatically adapts to new tasks and architectures without a pre-tuned recipe.
Darwin Skill — It autonomously optimizes agent skills through a validation-gated ratchet mechanism.
fal — Access 1,000+ generative models and high-performance serverless GPUs.
FreeToken — Runs 290B+ MoE models on consumer gaming PCs using elastic edge resources.
GEAR — Organizes recursive self-improvement through evolutionary search and reproducible evaluation.

All OpenEvolve alternatives