v1
latestOpenAPI 3.0.42026-08-04342835917.6 KBReturns a page of the requested confidence zone of an A-vs-B matchup — the prompts A clearly wins, the coin flips, or the prompts B clearly wins.
Each vote is weighted by the annotator's reliability; a prompt is placed by the Wilson score interval on A's win share and, within the requested zone, ordered most-confident first (coin flips by effective sample size). The same vote-level filters as the matrix query apply. Benchmarks without granular votes return an empty page.
Path parameters
The id of the benchmark.
Query parameters
The id of participant A (the reference side of the split).
The id of participant B.
Which confidence zone of the matchup to page through.
The 1-based page index.
The number of items per page.
Filter by country.
Filter by language.
Filter by gender.
Filter by age_bucket.
Filter by occupation.
Filter by tags.
Filter by participant_id.
Filter by leaderboard_id.
Filter by run_id.
Filter by prompt_identifier.
Filter by voted_at.
How to combine the field filters: "and" (default) requires every filter to match, "or" requires any of them to match.
Response
OK