reasoning benchmark
PaperBench Leaderboard
PaperBench evaluates AI agents on their ability to replicate research papers through code implementation, experimentation, and reproduction of scientific results.
Updated Sep 5, 2026
Models3
Model coverage3
MetricScore
EvidenceB
PaperBench Ranking
Higher score ranks better on this benchmark.
Sort by
Rank | Model | Score | Percentile | Participants | Evidence | Evaluated |
|---|
| Rank01 | ModelAC | Score93.00% | Percentile100.00% | Participants3 | EvidenceC | Evaluated |
| Rank02 | ModelMA | Score63.50% | Percentile50.00% | Participants3 | EvidenceC | Evaluated |
| Rank03 | ModelMI | Score52.60% | Percentile0.00% | Participants3 | EvidenceC | Evaluated |
PaperBench Highlights
The leading models and scores on this benchmark.
PaperBench Score Distribution
A closer view of the leading scores on this benchmark.
The Top AI Models for PaperBench
The first five results on this benchmark, with official price and output speed added where the model identity can be matched.
What is PaperBench?
What PaperBench measures and how its scores work.
PaperBench is a benchmark for evaluating AI agents on their ability to replicate research papers.
It measures performance on complex, multi-step workflows involving code implementation, experimentation, and reproducing scientific results from academic publications.
Scores are shown in ratio. This benchmark is not independently verified and has an evidence level of B.
- Family
- PaperBench
- Modality
- text
- Primary category
- reasoning
- Score direction
- higher
- LLMBoard eligible
- No
- Evaluation key
- overall
LLM Stats. Benchmark scores retain their original unit. Overall score eligibility is shown separately.
FAQ
Common questions about PaperBench.
Which model scores highest on PaperBench?
Qwen3.8 Max is currently ranked first with 93.00%.
What are the top three models on PaperBench?
The current leaders are Qwen3.8 Max (93.00%), Kimi K2.5 (63.50%), and MiniMax M3 (52.60%).
Which PaperBench model has the lowest official input price?
MiniMax M3 has the lowest matched official input price at $0.30 input / $1.2 output per 1M tokens.
Which models are fastest among PaperBench results?
The fastest matched records are Qwen3.8 Max (60.87 tok/s via DeepInfra) and MiniMax M3 (6.61 tok/s via MiniMax).
Does the highest score result prove overall model quality?
No. This benchmark measures one defined capability or task. The overall LLMBoard score uses a separate aggregation across eligible benchmark evidence.