Benchmark
LongBench
What is LongBench?
Nominal context length is not effective context length. Benchmark Radar tracks 1 reported scores for LongBench, cited by 3 source documents, with links to its…
- Released
- 2023-08-28
- Evaluates
- long context
- Openness
- unknown
- Importer
- Model reports
- Reported scores
- 1
Source provenance
- Original evidence https://github.com/THUDM/LongBench
LongBench code
LongBench results and reported scores
Model reports
| Model | Organization | Reported value | Reported | Evidence |
|---|---|---|---|---|
| Qwen3.5-397B-A17B | Qwen | 63.2 | 2026-02-16 | evidence |
Scores are partitioned by the source that reported them and are never merged into a single cross-source ranking, because the sources measure different things and say so.
Which sources cite LongBench?
- DeepSeek-V4 DeepSeek, 2026-06-25
- DeepSeek-V4-Pro DeepSeek, 2026-04-22
- Qwen3.5-397B-A17B Qwen, 2026-02-16