{"name":"benchmark","stage":"ga","description":"Package benchmark is one honest score for any model, on the tests everyone quotes.","operations":[{"operationId":"get_benchmark_catalog","method":"GET","href":"/v1/benchmark/catalog","summary":"Is the canonical benchmarks this arena reads — the id, title, axis, item count, upstream source and direction of each, with native marking the ones the standardized harness runs today."},{"operationId":"get_benchmark_claims","method":"GET","href":"/v1/benchmark/claims","summary":"Lists the effective claims the caller may read, each labelled with the org that made it and the user who recorded it."},{"operationId":"post_benchmark_claims","method":"POST","href":"/v1/benchmark/claims","summary":"Records claims for the caller's org: one to correct a number, many to import a leaderboard."},{"operationId":"get_benchmark_compare","method":"GET","href":"/v1/benchmark/compare","summary":"Is the ONLY valid arm-vs-arm test: it pairs the two models on the items BOTH completed, and answers rescue and damage counts with an exact-McNemar p."},{"operationId":"get_benchmark_history","method":"GET","href":"/v1/benchmark/history","summary":"Returns each model's measured score per run over time, oldest first, with the change between runs."},{"operationId":"get_benchmark_leaderboard","method":"GET","href":"/v1/benchmark/leaderboard","summary":"Answers one row per model for the benchmark named — what our own harness measured, beside what the vendor claims, and the gap between them."},{"operationId":"get_benchmark_presets","method":"GET","href":"/v1/benchmark/presets","summary":"Are the router blends available to compose from — a named set of model arms, the rank they escalate through and the panel width that bounds fan-out — each named for the model layer as enso-\u003cname\u003e."},{"operationId":"post_benchmark_presets","method":"POST","href":"/v1/benchmark/presets","summary":"Stores a router blend for the caller's org — its name, its arms, the rank they escalate through and the panel fan-out width — and answers with the stored blend and the enso-\u003cname\u003e it would be resolved under."},{"operationId":"post_benchmark_runs","method":"POST","href":"/v1/benchmark/runs","summary":"Answers what the arena already holds for a model on each benchmark asked about, and spends nothing."}],"_links":{"describedby":{"href":"/v1/openapi.json?capability=benchmark"},"self":{"href":"/v1/benchmark"},"up":{"href":"/v1"}}}