InfiniteBench

What it measures

Very long inputs test retrieval and understanding beyond ordinary short-context evaluations.

How to read the result

Inspect individual tasks and input lengths. Locating one item and reasoning over a long narrative or code input stress different capabilities.

Limits and comparison traps

Truncation, tokenization, and input construction can change what the model sees. Do not pool incompatible task metrics into an unexplained average.

Published evaluation

InfiniteBench aggregate; HELMET paper · original-author harness. The source’s task split, model configuration and scoring units are retained.

Where this fits in an evaluation suite

A large context window is a capacity specification. Successful retrieval and cross-document reasoning are observed capabilities. Test both at the actual lengths your workflow needs.

Start with a capability hypothesis, then choose a second test with a different failure mode. Keep the checkpoint, reasoning effort, agent tools, sampling policy, and scoring procedure fixed when measuring a training intervention. Record failures and missing outputs in the denominator.

A reproducible comparison

InfiniteBench aggregate; HELMET paper · original-author harness. The source’s task split, model configuration and scoring units are retained.

Publish the exact run configuration, item coverage, score units, and available uncertainty. If a source supplies only an aggregate or a single run, preserve that limit. A new model name or a rounded score alone does not establish a capability improvement.