What it measures

Stress-tests long-context output generation (generating 8k–64k token structured outputs, HTML, travel plans, and trace logs) rather than only long input reading.

Reading the measurement

Length is a test condition, not a capability score. Locate the relevant evidence, then examine whether the model resolves conflicts and combines details correctly. Keep document extraction, input length and answer grading fixed.

Published evaluation

Source average across procedural tasks at the named target-length difficulty. The length levels remain separate.

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

Source average across procedural tasks at the named target-length difficulty. The length levels remain separate.

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.

PRIMARY EVIDENCE & FURTHER READING

  1. LongProc: Benchmarking Long-Context Language Models on Long Procedural Generation — published result table ↗

    Xi Ye, Fangcong Yin, Yinghui He, Joie Zhang, Howard Yen, Tianyu Gao et al. · 2025

    All 8 authors

    Xi Ye, Fangcong Yin, Yinghui He, Joie Zhang, Howard Yen, Tianyu Gao, Greg Durrett, Danqi Chen

Reviewed 2026-10-26. Explanation and interpretation by RLTasks. Benchmark ownership remains with its original authors.