Skip to content
ReviewRamble

The weekly brief

AI, reviewed, in your inbox.

  • New reviews of tools, models, MCP servers and repos, with the verdict up front
  • Re-checked prices and ratings when something changes
  • New analysis pieces as they are published

Models, agents, MCP servers and the tools actually worth your time.

One email a week. Unsubscribe with the link in any brief. How we use your email

BrowseComp

BrowseComp is an OpenAI benchmark of 1,266 hard questions that tests whether a browsing agent can search the web persistently for tangled, hard-to-find facts, with short answers that are easy to check.

Also called: BrowseComp benchmark

By Divya Sharma · Checked Nov 15, 2026 · How we review

Many search benchmarks ask questions that a single query can answer. BrowseComp, released by OpenAI researchers in 2025, asks questions whose answers are buried: finding them may mean following many leads across many pages. The answer itself is short, such as a name or a date, so grading is simple even though the search is hard.

What it measures

The authors compare it to programming competitions for coding agents: incomplete, but useful. It tests persistence and creativity in finding information, the core skill behind deep research tools and AI search engines. It does not test writing long answers or dealing with vague user questions, which real research tasks also need.

Why it matters to builders

If your product relies on an agent that browses, BrowseComp scores tell you how well a model keeps going when the first searches fail. The paper also shows accuracy rising with browsing effort, which links it to test-time compute: more steps and more searches tend to help. Scores depend on the tools the agent has, so compare results run with the same search and browser agent setup.

Limits

The questions are designed so answers are easy to verify once found, which is not how most real queries work. Answers published online after release could also leak into model training, so check which version and date a reported score uses.

Related terms

Deep research (AI)
Deep research is an AI agent feature that plans a research question, runs many web searches, reads the sources it finds and writes a long report with citations, taking minutes rather than seconds.
AI benchmark
An AI benchmark is a fixed, public set of tasks with an agreed scoring method, used to compare models on the same test, for example coding issues solved, exam questions answered or retrieval quality across many datasets.
Browser agent
A browser agent is an AI agent that controls a web browser to complete tasks, such as searching, clicking, filling in forms and reading pages, often running a headless browser with no visible window.
AI search engine
An AI search engine is a search service that runs web searches for a query, reads the results with a language model and returns a written answer with links to its sources, instead of only a list of links.
Test-time compute
Test-time compute is the amount of computation a model spends while answering a request, rather than during training. Letting a model think longer or try several answers can improve results on hard problems.

Sources