Benchmark
OneMillionBench
What is OneMillionBench?
Long-context recall and use; a "with tools" figure is not comparable to a no-tools run. Benchmark Radar tracks 1 reported scores for OneMillionBench, cited by…
- Evaluates
- long context
- Openness
- unknown
- Importer
- Model reports
- Reported scores
- 1
Source provenance
- Original evidence https://github.com/humanlaya/OneMillion-Bench
OneMillionBench code
OneMillionBench results and reported scores
Model reports
| Model | Organization | Reported value | Reported | Evidence |
|---|---|---|---|---|
| Hy4 preview | Tencent | 65.4 | 2026-08-28 | 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 OneMillionBench?
- Hy4 preview Tencent, 2026-08-28