GDPval
GDPval measures AI's real-world economic performance across 44 occupations, inspired by Gross Domestic Product.
GDPval is a new benchmark, or test, for artificial intelligence. It's designed to see how well AI models can perform tasks that have economic value, similar to how Gross Domestic Product (GDP) measures a country's economic output.
The test focuses on specific professional jobs. For example, it might evaluate how well an AI can draft legal documents, analyze financial reports, or write code, tasks that contribute to the economy.
GDPval is an evaluation framework designed to assess the economic utility of AI models by measuring their performance on well-defined knowledge work tasks across 44 distinct occupations. It aims to quantify AI's potential contribution to economic productivity by benchmarking against human professional performance in these roles.
The benchmark draws inspiration from Gross Domestic Product, seeking to provide a comparable metric for AI's impact on economic output. It focuses on tasks that are both measurable and have clear economic implications, moving beyond traditional academic benchmarks.
This term is appearing now because AI labs are increasingly focused on demonstrating the real-world economic value and practical applicability of their advanced models. Benchmarks like GDPval are crucial for showing how AI can translate into tangible productivity gains and economic contributions, moving beyond theoretical capabilities.
- Introducing GPT-5.2OpenAI · OpenAI News ·
- Measuring the performance of our models on real-world tasksOpenAI · OpenAI News ·
- OpenAI ChatGPT
- OpenAI has introduced the GDPval benchmark to measure the economic performance of AI models across various professional tasks.
- Gross Domestic Product (GDP)
- GDP measures the total economic output of a countr
- MMLU
- MMLU is a broad benchmark testing knowledge across many subjects, whereas
GDPval was introduced by OpenAI as a new evaluation designed to track how well AI models perform on economically valuable, real-world tasks, drawing inspiration from Gross Domestic Product. Source ↗
Ask an AI assistant like ChatGPT to draft a simple business proposal for a new product. Then, ask it to summarize a complex news article into three bullet points. Observe how well it follows instructions and produces output that could be useful in a professional setting.
The Playbook: reusable how-tos, prompts, configs and tips pulled out of the news.
- PromptRole-play a student struggling with math for AI tutor simulationSimulate a student's math struggles and uncertainty for AI tutor evaluation. · 5 min
- TemplateReusable workflow templates for healthcare teamsTeams can create and share templates for common workflows like documentation or prior authorizations. · 15 min
- TemplateTemplate for building an AI tool tailored to your workA structured approach to defining and building a custom AI tool for specific business needs. · 15 min
- TemplateEvals CSV prompt list for agent skillsA CSV file template for listing prompts used in evaluating agent skills. · 20 min
- HeuristicDouble coding agent task completion with docs linksProviding coding agents with a documentation link significantly increases their task completion rate. · 2 min
Questions people ask
- What kind of tasks does GDPval test?
- GDPval tests well-defined knowledge work tasks across 44 occupations, such as drafting documents, analyzing data, or writing code, that have economic value.
- How is GDPval different from other AI tests?
- Unlike general knowledge tests, GDPval specifically focuses on tasks that have direct economic implications and aims to measure AI's contribution to productivity.
First seen on Subvolts Sep 25, 2025. Updated Sep 15, 2026. Written by an AI model from the items above and reviewed against them weekly; report an error.