Benchmark
PaperBench
PaperBench is a benchmark for evaluating AI agents on their ability to replicate research papers. It tests models on complex, multi-step workflows…
- Modality
- text
- Categories
- reasoning, agents, code
- Openness
- unknown
- Source
- llm_stats
- Reported scores
- 3
Reported scores
llm_stats
| Model | Organization | Reported value | Reported | Evidence |
|---|---|---|---|---|
| Kimi K2.5 | Moonshot AI | 0.635 | 2026-01-27 | evidence |
| MiniMax M3 | MiniMax | 0.526 | 2026-06-01 | evidence |
| Qwen3.8 Max | Qwen | 0.93 | 2026-08-02 | 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.