agents benchmark
MLE-Bench Leaderboard
MLE-Bench evaluates AI agents on machine learning engineering tasks by measuring their performance on Kaggle competitions.
Updated Sep 5, 2026
Models2
Model coverage2
MetricScore
EvidenceB
MLE-Bench Ranking
Higher score ranks better on this benchmark.
Sort by
Rank | Model | Score | Percentile | Participants | Evidence | Evaluated |
|---|
| Rank01 | ModelGO | Score63.90% | Percentile100.00% | Participants2 | EvidenceC | Evaluated |
| Rank02 | ModelGO | Score39.20% | Percentile0.00% | Participants2 | EvidenceC | Evaluated |
MLE-Bench Highlights
The leading models and scores on this benchmark.
MLE-Bench Score Distribution
A closer view of the leading scores on this benchmark.
The Top AI Models for MLE-Bench
The first five results on this benchmark, with official price and output speed added where the model identity can be matched.
What is MLE-Bench?
What MLE-Bench measures and how its scores work.
MLE-Bench is a benchmark for evaluating AI agents on machine learning engineering tasks.
It measures performance on Kaggle competitions using the Score metric, expressed as a ratio.
Scores are shown in ratio. This benchmark is not independently verified and has an evidence level of B.
- Family
- MLE-Bench
- Modality
- text
- Primary category
- agents
- 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 MLE-Bench.
Which model scores highest on MLE-Bench?
Gemini 3.6 Flash is currently ranked first with 63.90%.
What are the top three models on MLE-Bench?
The current leaders are Gemini 3.6 Flash (63.90%) and Gemini 3.5 Flash-Lite (39.20%).
Which MLE-Bench model has the lowest official input price?
Gemini 3.5 Flash-Lite has the lowest matched official input price at $0.30 input / $2.5 output per 1M tokens.
Which models are fastest among MLE-Bench results?
The fastest matched records are Gemini 3.6 Flash (62.56 tok/s via Google) and Gemini 3.5 Flash-Lite (21.45 tok/s via Google).
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.