What it measures
Evaluates whether LLMs can sustain coherent, high-quality generation over 2,000 to 20,000+ output words.
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
Published overall score combining output-length compliance and judged writing quality. LongWriter-6k is training data, rather than a scored test set.
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
Published overall score combining output-length compliance and judged writing quality. LongWriter-6k is training data, rather than a scored test set.
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
- LongWriter: Unleashing 10,000+ Word Generation from Long Context LLMs — published result table ↗
All 9 authors
Bai, Yushi, Zhang, Jiajie, Lv, Xin, Zheng, Linzhi, Zhu, Siqi, Hou, Lei, Dong, Yuxiao, Tang, Jie, Li, Juanzi
Reviewed 2026-10-27. Explanation and interpretation by RLTasks. Benchmark ownership remains with its original authors.