OpenAI Releases GABRIEL Toolkit to Transform Social Science Research

Ted Hisokawa Mar 03, 2026 20:48

OpenAI's new open-source GABRIEL toolkit uses GPT to convert qualitative text and images into quantitative data, enabling researchers to analyze millions of documents at scale.

OpenAI's Economic Research Team has released GABRIEL, an open-source Python toolkit that transforms qualitative data—text, images, interviews, social media posts—into measurable numbers that researchers can actually analyze. The toolkit, announced February 13, 2026, targets economists, social scientists, and data scientists who've long struggled with the impossible task of manually processing massive qualitative datasets.

Here's the core problem GABRIEL solves: qualitative data contains some of the richest insights about human behavior, but converting it into rigorous evidence has traditionally required armies of research assistants or simply gets abandoned as unfeasible. GABRIEL lets researchers describe what they want to measure in plain English—something like "how family-friendly is this job listing?"—and then applies that same question consistently across thousands or millions of documents, returning a numerical score for each.

The practical applications span disciplines. Researchers can analyze scientific paper collections to track methodological evolution over time. Educational researchers can measure how course curricula allocate attention across subjects. Historians can extract structured data from records covering every small town across Europe. Consumer researchers can identify patterns in what people actually value from review databases.

OpenAI's accompanying paper benchmarks GPT's accuracy at labeling qualitative data across multiple use cases, reporting high accuracy rates. Beyond basic measurement, the toolkit bundles several utilities researchers commonly need: merging datasets with mismatched columns, smart deduplication, passage coding, and deidentifying personal information to preserve privacy.

The toolkit requires minimal technical background, according to OpenAI, and ships with a tutorial notebook for getting started. OpenAI says it plans ongoing improvements based on academic community feedback.

For the broader AI development community, GABRIEL represents OpenAI's continued push into specialized research tooling beyond consumer-facing products. The release follows OpenAI's February 26 partnership with Pacific Northwest National Laboratory on federal permitting and its Figma collaboration on code-to-design workflows—signaling an aggressive expansion into enterprise and institutional applications during early 2026.

• openai
• gabriel
• ai research tools
• gpt
• social science