BrowseComp what it measures, how it is scored and who leads it

Name: BrowseComp
Creator: OpenAI
Published: 2025-04-16

Agents and tools · OpenAI · introduced Apr 16, 2025 · Accuracy

About this benchmark

Task: Find hard-to-locate, entangled facts by persistently browsing the web.
Dataset: 1,266 questions with short, verifiable answers.
Method: Short answer compared with the reference by a model grader.
Metric: Accuracy
Organization: OpenAI
Introduced: Apr 16, 2025
Paper: BrowseComp: A Simple Yet Challenging Benchmark for Browsing Agents (arXiv 2504.12516)
Home: github.com/openai/simple-evals

Newest first. New versions are added, never rewritten.

Version	Date	What changed
1Released with the paper.	Apr 16, 20251 year ago	Released with the paper.

Sources: each benchmark's paper (arXiv) and its official site or leaderboard. Scores are the boards captured here every day, exactly as published.