GEPA
GEPA is a framework that improves prompts, code, and agent settings by testing and rewriting them automatically for developers.
What it is good for
- Improving a system prompt with automated test runs
- Tuning an AI agent's instructions against a scoring metric
- Optimizing code or configuration text using evaluation feedback
- Reducing the cost of getting good results from cheaper models
At a glance
- Who it is for
- Setup difficulty
- Developer ยท Requires programming
Level 5 of 5. You need to write code to use it.
- Cost
- Free and open source
- Type
- Framework or SDK
- Tasks
- Works with
- DSPy, MLflow, Pydantic AI, LangChain, Claude Code, Comet ML Opik
Momentum
- Rank this week
- Not ranked yet
- Best rank
- None yet
- GitHub stars
- 7,443
GitHub numbers as of Nov 16, 2026. Updated weekly.
Official links
Setup steps are on the official pages. We do not repeat them here, since they change.
Questions
- Is GEPA free?
- Yes. It is free and open source (MIT license).
- Do I need to know how to code to use GEPA?
- Yes. You need to write code to use it.
- What does GEPA work with?
- According to its own documentation, it works with DSPy, MLflow, Pydantic AI, LangChain, Claude Code, and Comet ML Opik.
- What is GEPA good for?
- Improving a system prompt with automated test runs, tuning an AI agent's instructions against a scoring metric, optimizing code or configuration text using evaluation feedback, and reducing the cost of getting good results from cheaper models.
Similar agents to GEPA
- UAEF
UAEF is a framework for scoring AI agent quality across tool use, answer quality, safety, speed, and reasoning for developers testing agents.
- AssetOpsBench
AssetOpsBench is a framework and benchmark for building and testing AI agents that handle industrial equipment maintenance for researchers and engineers.
- exxperts
exxperts is a desktop app that keeps a reviewable memory for each project so AI answers build on what you confirmed.
- AgentDescent
AgentDescent is a Python framework that improves prompts, skills, and agent setups by testing them against your own examples for developers and builders.
- caveman
Caveman is a skill and proxy that cuts token use by making coding agents answer tersely and shrink the text they read, for developers and heavy AI users.
- career-ops
Career-ops is an open-source job search agent that scores job listings against your CV and drafts tailored applications for you to review and submit.