Reflective prompt optimization with DSPy GEPA
A cookbook published on October 16, 2025 presents reflective prompt optimization with DSPy GEPA and reports an 11% accuracy gain on NuminaMath-1.5 for less than US$0.50 in total.
Source: Otimização de prompts com DSPy GEPA (huggingface.co). Text prepared with AI from this source.
What happened and what to do
The cookbook, published on October 16, 2025, presents an example of reflective prompt optimization with DSPy GEPA. The method uses separate models for inference and reflection; on NuminaMath-1.5, the publication reports an 11% increase in accuracy for less than US$0.50 in total.
A company can apply this approach to its own tasks by first defining a representative evaluation set and a quality metric. A pipeline can test prompt variations, record results and costs, and track performance changes before a configuration goes into production.
How the consultancy can help
Wendelmaques can assess the opportunity and available evaluation data, design an optimization workflow, and implement its execution and quality and cost monitoring on the company’s infrastructure.
Next step
Send a short description of the task and expected outcome to receive a scoped proposal.
Consulting for your project
Infrastructure review, deployment and ongoing operations, with scope and pricing defined in the proposal.
Quoted per project
Request a proposal